How about an economic take to it- what if all civilizations proceed to a GINI index of 1? You could predict that increased automation and computation would result in less need for other people, so eventually one entity "wins" and consolidates all the resources of its home system.

After which, there's no spreading- it's not economically efficient to retrieve resources from other star systems, and there's probably no reason to migrate to another in a timeline shorter than hundreds of millions of years at least.

The universe might be filled with solipsists, or at least powerful beings trading maximally-compressed data...which looks like static to us.

The Fermi paradox does make it look like our understanding of physics and biology is logically incompatible, and so one will have to give.

One point to bring up is dark matter. If stuff that doesn't interact normally with matter exists, then philosophically it is plausible that so does double dark matter: something you could only detect once you had a full understanding of dark matter, and then spotted the one remaining anomaly. That chain could continue indefinitely, and so sooner or later you might come across a regime of interaction that had something like the properties of organic chemistry. If that domain was somehow drastically easier to 'reach' than other stars, expansion would proceed in that 'direction'. You can't even estimate the bandwidth, but it could easily be very high, meaning that a collapse of civilisation would propagate along it.

Cheap FTL travel could provide a similar means of making galactic colonisation impossible.

Or maybe biology is wrong. Perhaps the word 'organic' does actually mean what most non-chemists mean when they say it: 'magic'. Some kind of centimeter-scale quantum effect that means self-replicating probes with the properties of life plus accurate checksums are impossible to build.

Why can't I shake the feeling that Drake's equation is changed to include a new unknowable and presumably small factor whenever one of the old ones turns out to likely be orders of magnitudes larger than originally assumed? (Namely the amount of organic chemicals in the solar system and the fact that our solar system is probably not all that special in terms of the number of planets around your typical star.)

One point to bring up is dark matter. If stuff that doesn't interact normally with matter exists, then philosophically it is plausible that so does double dark matter

You misunderstand dark matter.

It's matter in the sense of being stable subatomic particles; it's just that they don't interact much, if at all, via strong, electromagnetic, or weak forces -- we only feel their effects via gravity. (For a long while, neutrinos were a strong candidate.) Your philosophical regression is therefore rather pointless.

As for biology being wrong, that's not terribly likely -- 18th century vitalism and mind/body dualism notwithstanding.

Note: it helps to be aware of the Great Filter theory (original paper here) before we dive into the Fermi paradox and the Drake equation. The key question being -- if the GF exists, have we met it already, or does it lie in our future? (The latter would probably be very, very bad news indeed.)

I disagree it looks "like our understanding of physics and biology is logically incompatible" - there's nothing in the paper or the ideas behind it that suggest anything like that.

What they do suggest is our understanding of behavioural psychology is still incredibly racist, in that it only considers a human set of motivations. Why should our advanced aliens want to colonise the galaxy, or even bother sending probes to find out what every system looks like?

Maybe space travel simply isn't interesting to other species. Perhaps they only bother when their local star is about to go nova and they need to decamp to the nearest safe planet (or find a way, however improbable it sounds, to sit it out in safety and carry on afterwards).

Maybe we should be surverying systems near to recent novas. At any rate, we can't assume any alien intelligence will place the same emphasis on exploration and colonisation as we (currenty) do.

What they do suggest is our understanding of behavioural psychology is still incredibly racist, in that it only considers a human set of motivations. Why should our advanced aliens want to colonise the galaxy, or even bother sending probes to find out what every system looks like?

Ahem: while I agree that there are implicit behavioural assumptions underlying the Drake Equation, I think you missed one angle -- which is that all it takes is for one alien ecosystem to express a curiosity-trait at the same time as the phenotypic adaptations to permit interstellar travel and, via percolation, it's likely to spread throughout the galaxy like kudzu. As we are such a system (minus the travel adaptation -- so far) we have to consider this as a plausible speculation.

I really don't like the DE's emphasis on prioritizing conscious/sapient civilizations and expansionary colonization/resource acquisition. But that's a different angle to take.

Ay, you're right of course - we're like roaches, it only takes one and the place is infested!

It may be we're just not looking hard enough. If there was a bus-sized probe sat on a small rock in the asteroid belt, periodically sending out a tight-beam report to a star within our galaxy, there's no guarantee we'd detect it any time soon. And if it's a nano-sized observer we may never detect it until we've been here a billion years,

One option is that (a) it's a nanoscale package, and (b) it's been here around 3.6 billion years. In fact, since the dawn of DNA-realm life on Earth. We have met the aliens and they is us ...

(File under "panspermia hypothesis".)

The paper is obviously nonsense. Compare our advances in telescopes with our advances in interstellar robotics. We'll have telescopic pictures of aliens tying their shoes before we've got an interstellar robot. At the very least, we'll know which nearby star systems have planets with cities, and we'll just go there.

At the very least, we'll know which nearby star systems have planets with cities, and we'll just go there.

You're betraying your assumptions. Did you read Karl's take on it and understand the implications of "any sufficiently advanced civilization is indistinguishable from nature"?

Consider: you're assuming we look at earthlike planets and examine them for signs of civilization -- city lights, for example.

But city lights are signs of an inefficient technology.

Go back 200 years and we didn't have them -- cities at night were dark and dangerous places, lit here and there by oil lamps and candles.

Go forward 100 years and it's possible that our much more advanced cities will again be dark, because you only need to light them up to allow people to navigate in the dark. If all vehicles are self-driving and equipped with radar (and designed to avoid ambulatory obstacles), they don't need headlights. And we've already got starlight vision goggles, albeit bulky and inconvenient; what if we shrink them to the size and convenience of spectacles, or are accompanied everywhere by small UAVs that shine bright LED spotlights on whatever we happen to look directly at?

I'll grant you that such "dark" cities may still have a microclimate effect (lots of organisms living in a dense clump metabolize more and put out more waste heat) but looking for a 1-5 celsius elevated heat signature on a patch of land a few kilometres across at interstellar distances is a very different proposition from looking at a city that's pouring hundreds of megawatts of light into the night sky.

Again: radio. The 1960s SETI program assumed that something like a cold war ballistic missile warning radar emitter, or a long wave transmitter designed to broadcast to a continent, might be visible across a few light years. But we've shut those down, for the most part. Our radio spectrum use is getting far more efficient ... and lower powered. Even military radars are falling quiet, because stealth technology means that if you emit, something invisible can sneak up on you.

Karl's speculation makes a lot of sense to me: advanced alien civilizations may be harder to detect than one at our level of development, in which case the window of detectability for an alien civilization may be less than two centuries (or equivalent, in terms of their innovation cycle speed).

Feel free to label me as a crackpot for saying this...it sounds like crackpottery to me...but I'm saying it anyway.

How about taking Karl's (very sensible) argument in another direction -- how do we know that what we observe in the Universe, and assume to be the result of natural forces...is entirely natural?

How do we know that what we are seeing is not influenced by intelligence? We're making assumptions about how advanced civilization behave, but what if their behavior looks like large scale natural processes to us. What if our expectations of what the result of natural processes look like is biased by unrecognized intelligent activity.

Charlie at #11 has made more sense than I ever would, but I was thinking as well that the issue is surely as much detectability by our current technology. If for example they are using lasers between systems, what are the chances of us actually noticing. And more far fetched, if they have worked out how many dimensions the universe actually has and how to use that to their advantage, not to mention exotic particles, then we couldn't detect anything at all.

Who knows, maybe pulsars are artificial?
(Although they don't seem to live very long and anyway in a universe where FTL is not possible I don't see why you would need them for anything. Anyway, just a thought)

Not a crackpot idea at all - what if the reason all galaxies are accelerating away from each other is the ultimate way to avoid conflict between high-end societies?

Although, pondering on Charlie's response #9, if Panspermia is anywhere ballpark correct (or Schroeder's), it might imply one good thing - super advanced beings are, if not benevolent, then at least are not outright malevolent. If they were, they'd have fragged us already.

(Or we're so far behind, we don't even register as a threat...Or they like to give other races a sporting chance before laying the smackdown...At any rate, it's good enough to reassure me we aren't up against Peter Watts-style ultra-vicious competitors)

You're betraying your assumptions. Did you read Karl's take on it and understand the implications of "any sufficiently advanced civilization is indistinguishable from nature"?

And thus Karl betrays his own assumptions. We don't know what a "sufficiently advanced civilization" looks like, (mine has lots of pretty lights) but there are lots of things which indicate the presence of civilizations, life, or water that could be seen through a telescope and point us at particular planets/solar system:

* Reflections from solar mirrors and panels. * Square/straight pieces of land - crops, gardens, etc. * Shapes not found in nature - pyramids, large buildings, etc. * Reflections from water with spectra that indicate industrial chemicals * Satellites around the target planet with a very high albedo (if inside the liquid water zone) * Infrared readings not justified by a planet's albedo * Lots of green on continents * Obvious arctic areas * Occasional explosions (war?) * Flames from space-drives. * Space elevator

We'll be able to spot all that and much, much more long before we can build our first interstellar probe.

But if we're going to believe in SRPs, why the very naive assumption that any civilization wants their SRPs to expand endlessly? Consider the following factors:

We don't want our probes to annoy the neighbors, who may be territorial, paranoid, greedy, or possessed of some other quality which means they would decide to make war on us once our probe started using their solar system for building materials. (One relativistic bomb can ruin your whole day.) This means that probes will be discreet and stealthy, and they won't eat too many resources the local civilization might want.

We don't want probes to expand beyond a distance where they could successfully send data. (Yes, I know probes could relay for each other, but at limited bandwidth/interstellar distances congestion problems multiply.)

We don't want probes to use all the available resources, thus taking them from us during our next step of expansion. (Absent really amazing nano, an interstellar probe would be very large and use many rare substances. Assuming stealth, we don't build laser farms, and the probe must be able to maneuver independently, which also works against a laser farm.)

If we must (sigh) assume SRPs, the instruction set probably looks something like this:

Go the next star system away from Earth. Make sure it does not have intelligent life. If so, report back to us then leave he area at once. (At the very least, we'll want to transmit the latest stealth designs to the probe and have it build those before we explore that system.)

If there is no intelligent life, plant your replication factory on a discreet place well away from the local star. Then begin more serious explorations. The replication factory makes five copies of the original probe, then pushes the asteroid it is based on into interstellar space so it can't be found by anyone, then acts as a router for its ancestors and descendants.

If you are more than 100 light years from Earth, do not replicate, stop and request instructions.

It seems to be a whole lot of work for no gain. What does this prove?

If the author were establishing the foundations for space fairing civilizations in a fictional work then I would applaud the level of detail that he undertook. But, as a scientific work? I dunno.

Even given the constraints of imagining life on this planet. One could hypothesis the existence of a creature who's traits could be well reasoned. Eating habits, habitat, mating rituals, life/ death cycles..... etc. But, does any of this make the imaginary creature anymore possible? Does any math predict species that we haven't encountered yet?

The mathematics in this work seems to lend a unwarranted sense of credibility to fancy.

Can I provide an explanation?

As background, right now I'm reading Rob Hopkins' The Transition Handbook. On page 70, he has this graph called "the petroleum interval in a historical context." It's a graph of energy use over a 4000 year interval, centered on 2000 CE. Obviously, the next 2000 years are extrapolated, but assuming the Peak Oil argument is correct, then global energy use is a brief, jagged, volcanic peak on an otherwise fairly flat plane.

Here's the explanation for the Fermi Paradox: Earth is normal, and it has a normal allotment of fossil fuels and radioactives. Fusion power is hard to achieve (to levels of unachievable), and nuclear power plants are difficult too. We've ridden two centuries of cheap energy, and (absent a miracle breakthrough) we're coming to the end of that period. Right now, we couldn't build a self-replicating space probe if we wanted to, let alone a self-replicating interstellar probe. Absent a miracle, we're not going to be able to build such a probe in the future either, because the amount of energy we can invest in space technology is inevitably going to decline towards zero.

That's an explanation for the Fermi Paradox: we're normal, and we don't have the resources to either settle the stars or even talk with them for more than a brief hail. Now, we're growing past the peak where we could theoretically have done so. We're never going to the stars, and we won't be talking to them either.

The galaxy could be littered with thousands of mature civilizations. Mature, in this case, means post-Peak and low energy sustainable, and each mature civilization is confined to a single planet. The Peak Energy civilization interval is so brief (1-2 centuries) that species going through the Peak can never even establish communication with another Peak civilization before settling down to a quiet maturity. After the energy peak, species lack the resources to run facilities like Arecibo and fall silent.

This is grim from an SF perspective, but it certainly fits our experience as a species.

I haven't read the whole paper, but the dismissal of Sagan and Newman's argument in Section 2.2 seems to me to brush under the rug a lot of really, really complicated fault tolerance issues.

His argument that computers nowadays undergo a lot of transient errors but survive is demonstrably wrong; there's lots of fault-tolerance research built out of observing computers running outside an atmosphere, and things go really bad all the time. To claim "bad checksum => force stillbirth" is enough, for example, fails the most basic test: if the cosmic ray "hits the checksumming chip", you're hosed.

Of course, you could claim "substitute checksumming for state-of-the-art fault tolerance", but I can also substitute 100 years of observed faults with 10^8 years of all combined faults of all SRPs, and the expected number of expressing mutations (that is, ones which would change the behavior of the machine) is still likely to be >1.

This is not to say that SRPs would not be used by civilizations, but I think Wiley was too cavalier as well.

How do we know that what we are seeing is not influenced by intelligence?

We don't, and that's an excellent point. For one thing there's Nick Bostrom's Simulation Argument. And for another thing, the late Robert Bradbury (who came up with the idea of Matrioshka brains) argued that we may have mistaken evidence of ETI for gross cosmological structures. (Unfortunately his website has disappeared and I can't find his original article on the subject.)

There will, of course, be other arguments along these lines. One I've heard (but not seen developed to the same depth as the previous two) is that the golden age of intelligent life in the universe is over -- it lasted from the instant of the Big Bang to T plus 10^-33 seconds afterwards, in other words for the duration of the cosmological inflationary epoch, during which the universe grew in volume by a factor of 10⁷⁸ -- we're just living in the cold, dead, vacuum left over after all the interesting stuff ran down/ran out of energy!

Heh. This is where sf writers really show their worth - this one's already been done by Wilson in his Spin trilogy (And he's not even the first, iirc.) Short form: any such beast launched in this particular galactic era will be exposed to an advanced ecology evolved from like machines inserted into the interstellar media eons ago.[1] The chances of its surviving to replicate in large numbers is slim to none in such an environment.

[1]In fact, Wilson had it that the original seeders had long since passed away and that any galactic civilization, such as it was, was simply the cumulative sum of these machines acting as permitted within the constraints of this ecology. A galactic society of V'Gers and Nomads if you will.

How about taking Karl's (very sensible) argument in another direction -- how do we know that what we observe in the Universe, and assume to be the result of natural forces...is entirely natural?

and:

Not a crackpot idea at all - what if the reason all galaxies are accelerating away from each other is the ultimate way to avoid conflict between high-end societies?

Again, already been done in sf. The use of the latter idea, the artificial nature of galactic recession, was used all the way back in the pulp era by Hamilton. And at least as recently as Baxter in his Xeelee stuff.

I think David Brin came up with that idea (of a probe ecology) first in 1986: Lungfish, which I saw in The River of Time. Of course, you can trace the idea back to Saberhagen's Berserkers...

It's my feeling that we are in the midst of the great filter. If we make it through the next century we'll probably be past it.

OTOH, we've already come within 30 seconds of failing to make it past. So I'm nervous about our prospects. I can't really say we've made it past the nuclear section, but currently it's looking like we've made it past the worst part of that aspect. Clearly lying ahead, however, are sections based around biology, computers, and nano-machinery (unless that gets combined with biology). I can't see the exact shape, but the danger of the nuclear section was constrained by the cost of participation. That won't be significant at the organization level for the biology section. I don't currently see what the existential threat arising from computers is, but it's clear that there will be one. Possibly arising from a combination of cyber-warfare and automation of industrial processes, but also possibly arising from robotic soldiers. (Berserkers? Probably not. Nobody's THAT silly...are they?) Anyway, that's a decade or so off. And I haven't much of a clue about nano-machinery, unless it's an aspect of the biology threat. The biology threat is also vague. I'm not clear whether it's customized plagues or something more specific. Say a virus that causes people infected to want to buy your product. (Sort of an advance on the disease of mice that causes them to act recklessly around cats, or the Coca-Cola logo that tends to make people reach for it as if it were a nipple. [I'm not sure that thing about the Coca-cola logo is true, but it's something I've heard claimed, and even if it isn't true, the "specific virus" I'm talking about could MAKE it true.])

Soru @ 2 Like it Remember, though that D matter is almost certainly real whilst D energy is hadwavium by the string theorists. Um.

Charlie @ 5 Yes. A future GF would be very bad news. H B Piper suspected the latter - he was very stri=ong on religious-nutters-wrecking-promising-civilisations, for instance. (see also Charles H @ 23) & @ 11 And how do you spot a civilisation that has gons inside a Dyson Sphere? Ditto Alex R. @ 15 Never mind a Kardashev II - HOW do you intend to detect it/them?

I think David Brin came up with that idea (of a probe ecology) first in 1986: Lungfish, which I saw in The River of Time. Of course, you can trace the idea back to Saberhagen's Berserkers...

The trope of The Machines Live On - and going about their own business - goes back at least as far as Campbell for the case of interstellar civilizations.[1] In fact, imho, the idea has a distinctly pulpish feel to it; I wouldn't be surprised to see a story from the 20's along these lines (I think I remember reading at least one such but my recollections aren't enough to be able to home in on it with Google.)

[1]Saturn's Children is arguably a specific instance of the general case.

Thanks for commenting on my paper. The purpose of any paper that responds to and addresses likely errors in previous papers is to advance the dialog on and understanding of the topic in question. For example, the I saw serious errors in the percolation and societal collapse theories and sought to clarify those subjects in the scientific literature so that their positions would not be assumed concrete and undisputed in future research and papers. The discussion of SRPs followed a similar suit in that I found the past arguments against them (refrainment and in efficacy) to be flawed.

Cheers!

Thanks for commenting on my paper. Please understand that I was working within the lengths limits of research publications. As such, some of my arguments were certainly somewhat terse, as you reasonably point out. I admit, the toss-off comment about checksums was fairly glib and not meant to be taken as one of the primary arguments of the paper. I apologize if that was not clear.

Please note that I wrap the whole discussion up with the "litmus test" later in the paper, and I really think it speaks more fully to my overarching point than any other single section of the paper: If we don't think a human research crew would genetically mutate into a human cancer (too weird to even conceive), then such an argument against SRPs is correspondingly weak.

Cheers!

I vote for a Great Filter in our near future. The Old Ones come and eat the minds of folks in civilizations.

As overall cognitive capacity increases in a civilization, the probability of attracting the unwanted attention of dead-gods from beyond space-time approaches 1. It's just as plausible an argument as Shroeder's "re-wilding" (which I also find an interesting idea, worthy of discussion).

Thanks for commenting on my paper. Your point is excellent...with respect to nearby stars. Resolving all but the closest stars, however, is hopeless. Recent work suggests we might even identify artificial night lights by their spectra without necessarily resolving any "images" from such planets, but again this applies mostly to nearby systems.

Space exploration always has, and always will, involve a combination of both passive observation and direct visitation, since each method offers advantages over the other. Some ETI are likely to pursue a similar approach.

Cheers!

The Great Filter has to be 100% effective. Just one civilization escaping it and the SRPs populate the galaxy/universe. For me, the paradox remains.

After the energy peak, species lack the resources to run facilities like Arecibo and fall silent.

I don't find your assumption that there is no post carbon fossil fuel future warranted. Solar energy (even just earth based) is more than adequate to fuel our future. That ensures continuing space access, leading to SPSs to power our future.

As to efficiency. If our species maximally exploits energy, the result will be maximal entropy. Thus our planet should be very detectable by sensors looking for high entropy sinks.

Solar energy (even just earth based) is more than adequate to fuel our future.

Petroleum isn't just a great energy source, it is a great energy storage system. How do you store all the solar power that gets collected in your scenario? Even for 1 week storage, the engineering required looks pretty impossible.

(Cribbed from http://physics.ucsd.edu/do-the-math/)

That's got good story potential- primordial civilizations notice that the continuing expansion of the universe is going to end the good times. They devise and implement a countermeasure- gravity- to slow the expansion. Opponents dislike the notion of keeping the universe small (fill in reason here)- they create dark energy to keep expansion continuing.
Fourteen billion years later, the struggle continues...

If you have energy (lots of it), you just create your energy storage mechanisms of choice. Hydrogen, more complex liquid fuels, batteries, heat, flywheels...

Once you get SPSs, then you have effectively 24/7 solar electric power deliverable anywhere on the planet.

Petroleum isn't just a great energy source, it is a great energy storage system. How do you store all the solar power that gets collected in your scenario?

I pondered answering Heteromeles, but didn't want to side-track the discussion on points which have been addressed already.

Since it's brought up, though, energy storage is far from impossible - but choosing the most economically efficient option for a particular application can be messy. Flammable liquid or gaseous chemicals have much to recommend them, and given an energy surplus we can make them if we need to.

Of course, an Arecibo dish just plugs into the main grid and might be powered by the closest nuclear plant. We have lots of uranium.

I think the article greatly advances discussions on the 'Fermi Paradox', and pulls various threads together into an updated form. I liked the figures which looked very reminiscent of 'Conways: Game of Life', which made me wonder if his simulations were done as a form of Cellular automata.

The only part left out would have been discussion of the deep time it took for an intelligent tool using civilization to arise on earth. The age of the Universe 13.75 Gyears with the Solar System forming 9.15 Gyears into that 13.75 (well over 50% the age of the universe). If at any point in the 4.6Gyears our Solar system has been around an expansion cycle swept through you would think evidence of it could be found (let alone in mega-structures left in our galaxy). The fact that we do not see such structures or remnants implies tighter limits to elements of the equation.

My personal speculation has led me to believe that there are no hard limits to prevent just one ETI from the SRP route, and since so far no evidence has been found perhaps the low probability element is intelligent tool using civs. Ignore 'rare earth' theories, instead its 'rare intelligence', with deep time required to see any arise. But then why should be we the first? Deep time also means time for some other in the galaxy to have swept through.

Ultimately it may come down to intelligent tool using civilizations are so spread apart in time and space that the only hope of seeing evidence would be to come along after a wave of SRP in your neighborhood.

Combine a low probability of coincidence in time & space for life, with a low probability of intelligence arising and you could effectively explain the 'Fermi Paradox'... The universe is huge, time is deep, however what we do not know are the odds for intelligence to arise. Sadly it's not a 'rich universe' if effectively we are the only ones we'll ever see...

I do find it interesting conjecture that all the interesting things happened in the first few seconds of the Universe... perhaps our era is so comparatively cold and dead that we're the rarity!

As far as paradoxes go, I always thought that Fermi's Paradox was incredibly weak. The earth is 4.54 billion years old and for 4.53999 billion of those years anyone visiting would have said, "not much to see here" or perhaps "mostly harmless" and then moved on. Even if their civilisation lasted long enough (i.e. was effectively immortal) and they were bored enough to revisit planets every 10,000 years just on the off-chance they got interesting the chances of actual aliens meeting us during our industrialised period is vanishingly small.

A probe that stayed in the solar system is a much more likely scenario, but even the most rudimentary camouflage would probably be enough for it to escape detection (stick it in the asteroid belt and make it look like an asteroid). I would think that unless the alien civilisation actually wants to contact us we would have no hope of knowing that they are there.

The first colonizing civilisation to arise in each galaxy colonises it and engineers it for longevity and efficiency ie something like Matrioska brains. Everyone else evolves in simulations, and they all seem to have the universe to themselves. The really ugly ones get deleted and the really nice ones are invited up for tea and biscuits.

When you look at this type of question, I think it's useful to look at what lessons you can learn from one scale down - our planet. While it's only one data point, it does at least have the benefit of the right level of complexity.

• Biggest success isn't us, it's the insects, or the plants if you want, or even the bacteria if you push it.
• Success isn't synonymous with intelligence, instead it's adaptability, simplicity and profligate reproduction.
• The successes don't really have more than limited mobility, instead they use circumstances/others to hitch a ride. Most fail and die, some succeed; that's enough.
• They tend to kept to the shadows, actively hiding and inhabiting crevices to avoid being noticed (particularly as a food source).
• Expansion is implicit in how they behave, rather than an explicit strategic plan.

As as far as us spreading across the universe, look first to see how we could adapt ourselves to be adaptable to different environments, and how we could easily spread our 'seed' widely (cf star wisps). The more complex the solution, the more likely it is to be wrong.

Ooh I like it, now how would we get their attention... move some stars around to give them the finger? Too anthropomorphic. The simulator's bound to have a few bugs perhaps we could crash it, and leave us to come up as root on reboot!

Rebooting the universe?

OK, so fezzes are cool, but how would we notice?

Dammit, I read a short story this year on almost this exact theme...A man was in a virtual room with a parrot or similar in a cage, and he was the keeper of the simulations, but then his boss turned up and demanded they be closed down, meanwhile one of the particularly nasty virtual races broke loose of its simulation and made a nuisance of itself, while a little girl with a teddy bear in another sim was saved from her miserable life...Anyone know which one I'm talking about?

We can argue back and forth about whether we can exploit nuclear power fast enough to make a difference (I suspect no, especially in the wake of Fukushima), and whether solar in space works, given how much trouble we're having implementing solar in deserts in a break-even fashion.

The bigger issue is still there. We're unlikely to have cheap energy much longer, and once energy supplies start becoming scarcer and more expensive, we're also going to lose the ability to send a probe to the stars.

A space program isn't 20 guys off in the desert building a Saturn V, using a solar plant to make the liquid oxygen and fabricate the aluminum rocket skin. It takes huge infrastructure to build even a rocket to the moon, even if one ignores all the bureaucratic problems with NASA, big Aerospace, etc. Most of our space missions, after all, are relatively tiny, disposable communications satellites, and they still cost millions to hundreds of millions per launch.

Once energy supplies start tightening, it's going to be even harder to field a big space mission, and there's going to be even less incentive to do so. But lets assume that someone wants to do it.

The next question is what to use to power an interstellar mission. If we could build a fusion rocket, we could also build a fusion power plant, so our energy issues would be solved. Fusion power is still 20 years away, as it has been for the last 50 years or so. Ditto storing large amounts of anti-matter for centuries. We're, what, 10-15 orders of magnitude away from doing that (microseconds to centuries and atoms to tonnes)? If we could store anti-matter indefinitely, we'd have the battery we need to make gigantic solar and wind systems work. There's no sign that we'll be able to do that soon, or ever.

If we strike those two options for interstellar drives, that leaves something like a laser sail (and the aim on the drive laser has to be atomically precise to hit a sail across light years), or pulse fission (like Project Orion). Assuming one of these two works, someone has to answer the question of why we want to divert so much of our scarce energy resources into sending a one-way mission to another star, when that energy is needed by billions of people on Earth. That's a hard question to answer.

To make the probe work, we also have to build a self-replicating technology that "lives off the land" to grow new copies of itself on asteroids, using either solar power or ship power. This system has to survive centuries of interstellar flight intact. Currently, we don't have a clue how to do this. Most organisms can't do it either, even given a benign ecosystem to grow in. Remember that such technology does not spring up in a cultural vacuum. If we could build such a probe, we wouldn't have many of the problems we currently face.

That's the point: there are massive technical challenges to building a society that could build an interstellar probe of any sort, let alone one that would want to do so. As I noted in the first post, all we have to do is assume that humans are normal, that our current problems are normal, and it follows that it's very likely no species will ever send out a star probe of any sort, because we're unlikely to ever surmount those problems.

Computer virus approach of inserting yourself into the boot sequence, or in our case what we want executed with root privilege. Upon reboot use our backdoor to scan their ports, find an opening and start exploring...
Or perhaps the code is buggy enough we can hack the simulator enough to reverse entropy, though that might get their attention fast

I think Robert Sawyer dealt with that in one of his books. He assumed that the way this got dealt with was by sending stars back in time, thus increasing the mass of the universe at an earlier period in time and slowing the expansion of the universe.

Hi Keith,

The problem is this: Our ability to learn, through telescopes (or the next advance on the telescope,) about the most useful place to aim an SRP is already far ahead of our ability to build an SRP. Assuming that our ability to build a telescope improves at the same rate as our ability to build a probe, the telescope will always be ahead - thus no SRPs.

You've also failed to note the arguments (in # 15 above) for building probes which only reproduce in limited numbers.

Specifically addressing Wiley's paper, I'd point out that the idea of a Peak Oil leading to a post-space sustainable society is actually nicer than McInnes' "Light Cage," (p.7) Simplifying the argument, the light cage occurs where societies choke themselves out through increasing density to point of collapse, and the only way they save themselves is by colonizing new systems faster than the old ones die. If they can't colonize fast enough, they all die. In previous posts, we've already pointed out that the dichotomy of infinite growth or quick death is both a false dichotomy and deeply embedded in discussions about our future. I'm just pointing out that our current models of a sustainable society with lower energy use seem to rule out star travel as well, and represent a third way out of this worn-out dichotomy of expand or perish.

My general critique of the biology in this paper is that it's quite basic. Real systems are much more complex, and typically the problems emerge from the complexities, rather than from the basic theories used here.

For example: let's talk about the percolation problem, otherwise known as how to get to the next star while maintaining your probe manufacturing base at the existing star. If it's harder to get to the next star, then the probe has a choice: expansion ends where it is, or the probe improves its design until it can get to the next star. If a probe improves, it is probably going to be easier for that probe to recolonize back in the direction it came (because going back is an easier challenge than the one it currently faces). The further they go and the more improvements they make, the more dangerous probes become both to the parent civilization and to their own parents. Worse, the probes know that it's easier to go back the way they came than to go forward. Do we want improved probes coming back at us? If not, then is it worth sending them out at all, knowing they are going to eventually fail when the challenges get too big?

Another big issue with probes is the death rate and the time it takes to get successful colonization of the next star. If it takes a century to reach Alpha Centauri, how many probes do we send out? It will take about 105 years to learn the results of even the first probe, and if we send out 1000 probes as insurance, we may saturate the Centauri system with SRPs without meaning to cause harm. Probes will always face this dilemma, of how many children to make and send, and they may occasionally flood target systems by accident. Biological systems don't particularly care about flooding systems: weedy species will make millions of seeds to insure that a few survive, which is why they can be real problems if too many of those seeds sprout. If we're trying to be ethical about SRP reproduction (and trying to keep from having the solar system over-run by hundreds of mutated SRPs coming back at us), we're going to be stuck with probes carefully taking millennia to colonize from one star to another, waiting to hear that a child didn't make it before sending another probe in that direction. That in itself would slow down expansion dramatically.

Third, another reason we might see predator SRPs was (apparently) overlooked: if you're building an SRP, it's easier to cannibalize an existing SRP than to create one out of raw materials. That's why predation arose on Earth. I suspect that predation (or piracy) is one modification SRPs would figure out on their own, especially if they are allowed to improve their own designs. Thus predators may naturally arise if SRPs are sent out. They may come back at us, too. As I noted above, self-modifying SRPs become more dangerous to us and to their parents, the further they go.

Some other random comments: invasive species (weeds and pests) share many similarities with cancer, with the advantage that there's literature on how they colonize new lands. While I realize that the old literature used cancer as a metaphor for colonization, I'd suggest that becoming familiar with weed ecology might be more useful. Also, it's worth understanding a bit more about colonization events (see Carlquist's Island Biology for a primer on how organisms colonize islands. Some of it is applicable.).

"If we're trying to be ethical about SRP reproduction (and trying to keep from having the solar system over-run by hundreds of mutated SRPs coming back at us), we're going to be stuck with probes carefully taking millennia to colonize from one star to another, waiting to hear that a child didn't make it before sending another probe in that direction. That in itself would slow down expansion dramatically."

So where are the SRPs from civilizations that didn't have those ethical concerns? And if the probes can think about how many "offspring" to make...wouldn't there be a distinct selective bias for those that made them as fast as practical, rather than those that artificially limited their potential reproduction...so we're more likely to encounter those?

Biological systems are always more complex. But you need a model which is simple enough to use. Getting the balance right, to get useful results, is hard.

Look at the drivel from the early days of ecology as example. But you need to start somewhere.

Plausible solutions:

1: The elders have decreed that lifebearing planets and their solar systems are to be left the frack alone. Noone argues with their conservationist tendencies, because they planted nova bombs inside every single star in the galaxy during the 487 million years they were the only civilization around. - This answers the "why does everyone obey the prime directive" problem neatly - anyone that doesnt go exinct via exploding star/planet/dropping dead when the killswitch inserted into the primordial ooze they evolved from is activated. You cannot fight an enemy that got to spend millions of years rigging the battlefield. So dont make enemies of entities like that.

2: The galaxy is completely overrun with advanced civilizations. None of which have any particular interest in rocks floating in space - they all settle the turf where there there is plentiful energy and matter to play with- the stars. Yes, inside them.

3: Miniturization is the most effective way to make an interstellar probe both cheaper and faster. Taken to its logical conclusion, again, they could be here, and we really would not notice, because they are rather.. smaller. than we expect an alien probe to be.

Energy scaricity is a nonsensical answer, because nuclear fission breeders is an existance proof that technological civilization has buildable energy sources that will not run out, and not all societies are going to be radioactivity-phobic. Not even all human societies are! And if the French inherit the world because everyone else runs out of energy, that will be hilarious. Aliens need not even be vulnerable to radiation - a species that evolved on a particularily "hot" planet for whatever reason (Stuck in a gas gigiant radiation belt, ect..) could have Deinococcus radiodurans levels of resistance to radioactivity and sprinkle nuclear waste on food as spices.

As I understand it, thermodynamics means an advanced civilization can't be totally indistinguishable from its surroundings- the things they do will make at least some heat.

Which I guess leads to the question- what's the theoretical maximum efficiency for computation, and where's the best place in the universe to do it?

There seem to be two probable (but mutually exclusive) conclusions based on the discussion so far:

1) Aliens are here, there and everywhere, but we can't detect them with our current methods.

2) Space-faring races of any flavour are vanishingly rare across the entire universe.

Solutions in the middle ground seem to rely on all the aliens conforming to a single set of behavioural/technological limits, which seems inherently improbable.

Actually, reading that back I see that (1) also implies a universal non-malevolence behavioural trait, in light of our continued and unmolested existence.

Ergo parsimony tends towards (2).

Hey Alex, I don't understand the argument that we should logically forgo all future probe-based exploration just because our telescopes pretty darn nifty. After all, we're sending exploration probes throughout our own solar system hand over foot despite the fact that we could ostensibly just point telescopes at our own planets instead. Wouldn't the same logic apply? You're correct that telescopes are providing us with copious and valuable data about extrasolar planets, but that's a far cry from saying that we now understand everything there is to know about those planets, that no new information could possibly be gained by literally looking closer. The resolving power of our telescopes will always be intrinsically limited some degree. Thus, a probe sent much nearer a system of interest will always offer additional data beyond a telescope forced to operate from a tremendous distance...and that is obviously the justification for a probe-based exploration mission in the first place: that it offers data returns that remote observation cannot.

As to post #15, I find its assumptions about the motives and behavior of ETI to be extremely specific. I don't think we can assume that every independently evolved alien species and civilization in the galaxy will adopt the methods of exploration proposed in post #15. They are certainly plausible scenarios and some ETI might behave in exactly the ways you suggest; I just don't think we can confidently conclude that they will all necessarily behave that way.

If some civilizations prefer to only explore a 100ly radius, as you suggest, then that's fine, SRP arguments on the Fermi Paradox simply don't apply to those civilizations. It's the other civilizations the argument applies to, the ones that do initiate a pan-galactic exploration mission. Given the presumed diversity of ETI, and given that rampant exploration and colonization are perfectly feasible cultural constructions (as frequently demonstrated throughout human history), the Fermi Paradox then becomes a discussion about those particular civilizations, the ones driven to explore and colonize. Civilizations which are posited to stay at home or to not explore very far are admittedly off the table, but they aren't the ones up for discussion in the first place, so they don't seem relevant to me.

Also, bear in mind that even if a civilization only explores some finite distance, the frontier of the civilization might very well be under constant expansion. That is to say, as the civilization colonizes the galaxy, that 100ly radius is constantly pushed outward; it is, in effect, simply a 100ly buffer surrounding the expanding civilization. Eventually, such an approach would fill the galaxy. Hence arises the Fermi Paradox.

Cheers!

Dirk @ 38 Which explains the "Great Filter" too! Nice one - though perhaps not for us simulacra.

heteromeles @ 47 Tiptree "A momentary teste of being" ( Nasty, very nasty )

C @ 51 Dyson Spheres & similar - very difficult to spot

Perhaps the earth is a special place after all?

One remarkable feature of our solar system is that the local interstellar medium is particularly rarefied -- more so than the average for the galaxy.

Why would this matter? Perhaps the practicable method of interstellar travel involves some variant on the Bussard ramscoop -- even if just used as a braking system.

Galaxy might be buzzing with SRP's, but we don't see them because they can't stop here.

Of course, the Sun has only been passing through this locally rarefied region in its orbit of the galaxy for some tens of millions of years. So why isn't there an ecosystem of evolved SRPs buzzing around the Solar system -- the descendants of probes that were trapped here when the Sun moved out of regions congenial to interstellar travel?

The non-malevolence need not be voluntary, and is in fact unlikely to be so - Given an interstellar community of civilizations, a certain minium of common social norms governing interaction and general conduct is not only nigh-certain to exist, it is likely to have enforcement mechanisms. So that objection is actually quite easily answered. This of course, also implies that humanity will not get to villy-nilly colonize the planets of the galaxy - any enforcement mechanism that kept our ancestors safe all the way back to the primordial slime is not going to keel over in the face of "human gumption" or any other such nonsense. Want to do stuff outside native solar system? File the paperwork, and dont fuck it up.

Petroleum isn't just a great energy source, it is a great energy storage system. How do you store all the solar power that gets collected in your scenario? Even for 1 week storage, the engineering required looks pretty impossible.

Ever heard of the Fischer-Tropsch process?

We will run out of fossil fuel sooner or later but that doesn't mean we won't be able to use long chain hydrocarbons for our own purposes. Creating them will be energetically inefficient, but if the carbon source is CO2 extracted from the atmosphere its net environmental impact will be down to direct heat pollution only (no greenhouse emissions because net CO2 emission will be balanced by net CO2 consumption).

What will happen in this scenario is that the price of energy for nighttime processes, i.e. stuff that has to run 24x7, will climb steeply. But even then, this only happens if we deliberately avoid certain other pathways (lots of nuclear fission reactors, possibly thorium cycle reactors; no development of orbital power satellites: no development of long-haul intercontinental power grids: no development of flow battery technology).

If we could build such a probe, we wouldn't have many of the problems we currently face.

Turn the paradigm on its head: the research projects necessary to build such a probe will, in turn, solve a lot of our problems!

(Unfortunately the historical precedents suggest they'll bring new ones, but that's another issue entirely.)

Which I guess leads to the question- what's the theoretical maximum efficiency for computation, and where's the best place in the universe to do it?

You need to read Seth Lloyd's work. Short answer: the event horizon of a black hole.

We don't need exotic tech like antimatter to store energy from solar. We just need electricity, CO2 and water to octane. Plus, the cost of solar power is still decreasing, and there's no reason to suppose it won't effectively hit zero for the panels themselves. The final cost will be infrastructure.

"Ooh I like it, now how would we get their attention... move some stars around to give them the finger? Too anthropomorphic. The simulator's bound to have a few bugs perhaps we could crash it, and leave us to come up as root on reboot!"

I would suggest that it has no bugs. After all, it's probably been online for several billion years. As for getting their attention - we already have it. We are talking about Gods. As for what the trigger might be for the unveiling (or deleting), it may well be our creating such simulations ourselves, or possibly creating artilects. Our bit of the real universe would likely be far more computing resource poor than it appears.

You misunderstand dark matter.

It's matter in the sense of being stable subatomic particles; it's just that they don't interact much, if at all, via strong, electromagnetic, or weak forces -- we only feel their effects via gravity.

I don't see any misunderstanding. It's just a physicist's working assumption, by way of Occam's Razor, that dark matter only interacts with other dark matter by gravity. It could interact with itself through some of the standard forces, and 5 more unknown ones. The only effects would be on the distribution of dark matter at a smaller scale than we can currently resolve.

Occam's razor is all very well when things are consistent, but the Fermi paradox shows that the standard models of biology and physics are not. So something has to give.

That something could be panspermia, dark chemistry, FTL, vitalism, ultra-pessimism, Cthulhu, creationism, simulationism or solipsism.

Or it could be something really weird. Like a galactic empire or federation that has declared the Earth a nature reserve. Or 99% of the universe being asteroid-scale primordial black holes, such that the chance of a planet not being hit by one over a billion years is around 1/number of stars in sky.

Ugh. Suggests not quite a Lovecraftian Horror scenario, but close to it:
Alien civilizations do exist, they use galactic-center black holes for their computing, we can't see the one in the Milky Way well because the rest of the galaxy's in the way, and there aren't any other middling-advanced civilizations because every so often the really advanced one tosses a star or two into the event horizon to fuel some advanced calculation or just reload the hole and the resulting radiation sterilizes most of the galaxy.

No, not plausible. (The event horizon is the constraint on density and switching speed; if you go inside it, you've got no way of getting information back out. So the ultimate computer is very hot, very dense ... but not quite inside its Schwartzchild radius. Oh, and it's also compact. One big black hole is less efficient than lots of tiny ones.)

Alien civilizations do exist, they use galactic-center black holes for their computing, we can't see the one in the Milky Way well because the rest of the galaxy's in the way...

Hmm? We can "see" the Milky Way's central black hole better than any of the others in other galaxies. Current radio observations resolve emission on something very close to the scale of the event horizon (see here, for example).

If a probe improves, it is probably going to be easier for that probe to recolonize back in the direction it came (because going back is an easier challenge than the one it currently faces). ... the probes know that it's easier to go back the way they came than to go forward.

I'm not sure how you can argue that. It's not like a probe has to build a road through interstellar space to get to another star... (And, of course, if a probe happens to start off travelling in the direction of higher stellar density, it's more likely that "going forward" will mean a shorter distance to the next star than going back would.)

Another possibility - anthropic selection in the multiverse. The first civilisation to arise in each galaxy spreads through it and suppresses the development of intelligent life, either deliberately or as a side effect of its activities. So in the multiverse we are only ever in the position of being the first. Everywhere else we did not evolve or got wiped out early on. Probably just coincidence, but Homo Sapiens is so new that the lifetime of our species approximates the time that it would take to saturate the galaxy in machinery.

Remember, though that D matter is almost certainly real whilst D energy is hadwavium by the string theorists.

No, "dark energy" in general is phenomenologically motivated: the universe is (now) expanding too fast given the amount of normal (baryonic) and dark matter that's present. The simplest explanation that's compatible with General Relativity is some smoothly distributed form of energy with negative pressure. The simplest version of that is a cosmological constant: same density everywhere, with pressure exactly -1 times the energy density, and no changes with time. (The zero-point vacuum energy hypothesized by quantum field theory has the right properties, but the standard attempts to estimate what it ought to be end up 120 orders of magnitude too large.)

Specific proposals for the nature of dark energy may indeed be handwavium, but the current accelerating expansion of the universe is fairly well established from an observational point of view.

Admittedly, I'm toward the edge of my knowledge of astrophysics here, but don't larger black holes last longer? If someone's interested in having a computation source for a very, very, very long time, would there be an advantage of one large one with a galaxy of mass around it over many small ones?

if you go inside it, you've got no way of getting information back out. So the ultimate computer is very hot, very dense ... but not quite inside its Schwartzchild radius.

Possibly not:

Black Hole Information Paradox

Hi Keith,

You're missing my point. SRPs are a good strategy if you don't know much about the places you wish to probe. What an advanced telescope gives you is the ability to know in advance where you wish to send an ordinary probe which does not need to reproduce and thus can be much less complex.

In other words, a really good telescope plus a cheap, one-shot, non-reproducing probe is much cheaper (and far less risky) than designing and deploying a reproducing probe.

Consider the expensive, complex issues of designing, prototyping, testing, and building a reproducing probe. It is expensive to do so, and the testing phase could be particularly risky if the probe produces mutant offspring.

As to your statement, That is to say, as the civilization colonizes the galaxy, that 100ly radius is constantly pushed outward; it is, in effect, simply a 100ly buffer surrounding the expanding civilization. Eventually, such an approach would fill the galaxy. Hence arises the Fermi Paradox.

You've missed the other issues that arise out of # 15's "gentle" approach to SRP exploration. The probes would be stealthed far beyond our ability to detect them (we're talking about a really advanced civilization here) and they would behave in a stealthy manner - reproducing in the outer solar system, not coming close to an inhabited planet, avoiding our own space probes, and phoning home for instructions before approaching an inhabited planet. Furthermore, they would not reproduce to the point of being common.(Any civilization capable of building probes would be capable of building a mission profile that does not result in multiple probes visiting the same solar system.)

So the our failure to find any probes doesn't lead to a meaningful method of counting aliens. We can expect probes (SRPs or otherwise) to be stealthed, stealthily behaved, rare, and careful. Assuming light speed limits, any probes in our solar system are probably still awaiting instructions from home on what they should do with the newly discovered race.

Assuming light speed limits, any probes in our solar system are probably still awaiting instructions from home on what they should do with the newly discovered race.

That arrangement would not be useful. If a probe has to send signals to some galactic civilization at the core, the time to receive a signal and get a reply would be far too long to handle the rapid technological development of the observed civilization. Even a few hundred years might be far too slow to be useful.

It is far more likely that probes are autonomous, or that the communication mechanism is FTL.

Having just skimmed Seth Lloyd's 2000 Nature paper (I wasn't aware of it till Charlie mentioned Lloyd's work; a preprint version is here) -- Charlie is talking about black holes as the limit for computation, not as computational devices themselves. The idea is that you can get more done, faster, if you compress matter: more bits in a smaller space (one upper limit to computational speed is the time it takes light to travel from one side of the computer to the other). But you can't compress matter indefinitely: if you push the density too high, your computer collapses into a black hole.

Now, I did notice a small side discussion in Lloyd's paper on the question of whether you could use black holes for computation. If information (originally in the form of infalling matter and energy) is not truly destroyed in a black hole, but instead comes back out later in altered form via a black hole's Hawking radiation, then perhaps you could do computations with black holes. But this requires carefully controlling the creation of the black hole itself, and then monitoring all the Hawking radiation that comes back out as it evaporates. (And controlling or preventing the infall of other matter and energy, including the cosmic background radiation.) A black hole with the mass of the Sun requires 10^67 years to evaporate; the Milky Way's central black hole (4 million solar masses) would require 10^80 years. Rather a long time to wait for the answer.

One thing bothers me about "stealthy" probes. Our response could become paranoid about their invisible [God-like and potentially wrathful] presence. There is certainly an analogy with SETI's thinking that signals are not being broadcast like light beacons, but are rather much more transient and focused on target stars. As lack of evidence on ETIs is maintained, SETI spends aver more effort looking for less and less obvious signals. This behavior is not unlike the witch finders looking for evidence of Satan, destroying human bodies in their zealous search for those hidden "number of the beast" marks.

Although we haven't looked yet, suppose we find no evidence of [macro] probes in our solar system. Then the suspicion is that they are so small as to be invisible. If we rule that out, then someone will say that maybe they are just outside the solar system, maybe in the Oort cloud. And if not there, then maybe they are on the way. [I include simulation arguments in this category].

At some point perhaps we should simply accept that we really are alone, and allow the FP to collapse to its logical conclusion.

Or there is a nearby base to which probes can direct their request for instructions.

The probes would be stealthed far beyond our ability to detect them (we're talking about a really advanced civilization here) and they would behave in a stealthy manner - reproducing in the outer solar system, not coming close to an inhabited planet, avoiding our own space probes, and phoning home for instructions before approaching an inhabited planet.

I'm not convinced. Reason: energy density.

The energy cost of interstellar travel is high, even if we max out at sub-relativistic speeds (e.g. 1% of c). In the absence of a viable design for a Bussard ramjet that works as anything other than a brake, it's reasonable to assume that a newborn SRP must accumulate the energy and reaction mass for its onward voyage in the solar system where it is created. We can reasonably ignore chemical energy sources (the energy density is far too low for timely interstellar flight), which leaves us with: generate your own power through fission or fusion, or rely on solar power (either beamed from solar collectors, or accumulated and converted into antimatter for storage).

Running nuclear power plants in the outer solar system are going to stick out like a sore thumb thanks to their infrared emissions. While for beamed power or antimatter production from solar sources the SRP is going to ideally want to get as close to the star as possible.

The whole thing becomes greatly complicated if it turns out the neutrinos are FTL. I'm willing to wait on that one, and talk later...

I've made exactly that comment to Seth Shostak at a SETI meeting, albeit in a different context.

I tend to think that advanced probes will be at the very least, semi-intelligent, rather like the original version of the monolith in Clarke's Odyssey novels. Clarke did not assume the monoliths would be self replicating or capable of interstellar flight, so they were rare.

I also think we have to be less binary on the observation vs intervention of probes. The time scales we are talking about are vast. Perhaps self replicating probes do reach every star, determine if there is life, and if not seed the planets/stars with the appropriate minimal life form and allow evolution to take over. The probes still make rare appearances in our solar system, but degrade or leave once they confirm that life is established.

None of this explains the absence of cosmic scale engineering.

You've got a good point if we allow certain assumptions:

Some human astronomer is looking for infrared sources in the Kuiper belt or Oort cloud, and they are looking in the right place. (I've got no idea of the probabilities.)

The SRP factory is not off the plane of the ecliptic and somewhat further away. We might never notice an SRP factory in orbit around a brown dwarf eleven light-months away, or we might even mistake an SRP factory for a brown dwarf.

There is an SRP factory currently in operation nearby. (SRPs would be relatively rare given the disadvantages to unlimited reproduction.)

A highly advanced technology doesn't have some way to either hide the heat signature of a fission/fusion device, or generate power without heat. Remember that heat is evidence of an inefficient process.

The SRP factory is on the surface of some rocky or icy body. It could be inside a rocky/icy body, Sedna, for example, which would result in heat rising to the surface very slowly. A human astronomer who observed the planetoid wouldn't say "Oh God, I've found aliens." The response would be more like, "What an interesting ice volcano" or "why does this planetoid have a molten core?"

Hi Dirk,

Yes and no. If you're creating a model, the point is not to start with something simple, it's to make the model simple enough to generate useful insights. It needs to be an intelligible cartoon of reality. The biological concepts in this model appear to me to be a bit too simple, and as a result, I think the model may be misleading.

With respect to biology and ecology, the problem is not with the fundamental rules (which are often based on chemistry and physics), it's with the fact that many of them apply simultaneously. It's the interactions among these rules that generate the unforeseen consequences and critical limits. That's what's that's missing in Wiley's analysis.

For a gamers discussion of difficulties: http://www.projectrho.com/rocket/spacewardetect.php

You could hide quite massive robotic outposts from our current observational capabilities of the solar system if they have been here since before we developed telescopes ect. Heck, you could hide them on earth if they have been here long enough, and have dug in. An outpost built into the bottom of the canadian shield, and running of geothermal would not be something we could find very easily at all. Their observational gear on the surface otoh... hm. Try electro magnetic pulsing/irradiating a few interesting locations and subject the dust to highpowered microscopes.

In other words, a really good telescope plus a cheap, one-shot, non-reproducing probe is much cheaper (and far less risky) than designing and deploying a reproducing probe.

If you're only interested in investigating one system, that's probably true. What if you want to investigate 100,000 systems, or ten million?

It's not at all clear that the expense of researching, developing, and building a reproducing probe is that much more than that of a one-shot probe. (Especially since you want your "cheap, one-shot" probe capable enough to properly investigate the target system and report back to you.) And once you start looking at making many probes, the advantages of offloading the energy and resource consumption of building (and possibly fuelling) them onto other star systems starts looking more attractive.

Re the Great Filter argument:

Isn't it fairly surprising that Earth has been home to complex multicellular animals for around 500 million years and in all that time we are the only tool-using, adaptive-technology developing (TUATD) creatures to have evolved[1].

Given that it only took us ~10 000 years to go from being just another tool-using ape to being what we are now, impacting our environment on a global scale and all that stuff, it strikes me as surprising it hasn't happened on Earth at least once before.

So if the probability of TUATD creatures developing is 10,000/500,000,000 = 0.00002, that might be one big filter. Given half a billion years and it's only happened once...

Still we've had (in theory) a good few billion years for it to happen.

[1]: ISTR reading somewhere that if we all died off tomorrow evidence of our existence would be minimal within some relatively short period of time, in terms of deep time. This suggests that tool-using, adaptive-technology developing creatures might have existed on Earth prior to ourselves[2], but managed to kill themselves off/otherwise go extinct. If it is the case that TUATD creatures have developed several times in the past 500 million years then that suggests we have yet to meet the great filter.

[2]: But the oil and coal is still here, and that mostly formed around 300 MYA. So IF TUATD inevitably use and discover oil and coal as fuel sources (highly likely IMO) then this suggests that there haven't been any TUATDs at our level of development in the history of complex multicellular life on Earth. Maybe.

Since even octopi move rocks around, I'd suggest tool use is not a strong category.

As you noted, we have deep deposits (dating back to the Carboniferous) of coal, and iron deposits that were laid down over a billion years ago. These went unmined.

The evidence for a civilization of our scale would be hard to cover up, since basically we took a big chunk of unrefined ores of almost everything and threw it into the biosphere, along with a flush of chemicals (especially industrial polymers) which can't be degraded yet (if ever).

Given that the carboniferous basically happened when plants temporarily outran fungi by producing massive amounts of non-decomposable polymers (aka wood) and the traces were still around hundreds of millions of years later, I suspect that the fossil evidence of our period will linger for a very long time, unless we clean it up ourselves. Future sapients may call our period the Plasticene (or its linguistic equivalent).

Peter Erwin @ 69 The zero-point vacuum energy hypothesized by quantum field theory has the right properties, but the standard attempts to estimate what it ought to be end up 120 orders of magnitude too large. Which should tell you RIGHT AWAY, that it's seriously worng, somewhere, no?

IF TUATD inevitably use and discover oil and coal as fuel sources (highly likely IMO)

Not really; you're taking fossil fuel deposition for granted. In fact, the Carboniferous era is a real headache. It lasted around 60 million years, and the reason we have all this coal and oil is a huge coincidence: low sea levels coincided with woody plants developing lignin, an insoluble polymer fibre used by plants for structural support and armour (in bark). Nothing could digest the stuff, for several million years! So it just piled up in the soil, and ended up forming deep drifts of sedimentary material (and, ultimately, rock deposits).

There's no obvious reason why a structural polymer occupying the same role as lignin has to be so profoundly indigestible or give rise to such huge deposits; it may be the case that biospheres with multicellular life seldom end up in such an inefficient carbon sequestration cycle, in which case intelligent tool-users rarely have fossil fuels to work with.

In fact, if in general tool-using adaptive-technology developing species have to leap straight from the equivalent of wood-burning tech to nuclear fission, that would be one candidate for the Great Filter: we're well past it, but we're very unusual.

So if the probability of TUATD creatures developing is 10,000/500,000,000 = 0.00002, that might be one big filter. Given half a billion years and it's only happened once...

So assuming a universe of 14bn years old, and a galaxy of 400 bn stars, that would be:

14/0.5 * 4E8 > 1E10 civilizations since the age of the galaxy. That would be a LOT of contemporaneous civilizations. Back to the Drake eqn...

Probably not a good idea to base probabilities on a sample of one.

The entire concept of 'stealthily probes' runs up against thermodynamics. Unless the probes are already in place and virtually inactive they are going to shed detectable waste heat. The act of transmitting signals back home will raise their temp above ambient, as will doing autonomous computation or maneuvering.

Beating your argument without handwavium will be very difficult indeed. Let's imagine some scenarios:

Case I.) This is the scenario we currently find ourselves in, so I'll start here. That is, Humans inhabit a planet which uses a petroleum-based, low tech (to the probe) energy intensive, highly detectable technology. The fact that Earth is inhabited by smart critters can probably be detected from 20 light-years away, and this has been true for at least 50 years, based on radio waves alone, without reference to heat signatures, atmospheric analysis, etc. It is true that radio waves don't do a good job of getting off the planet in coherent form, but the point-source orbiting 93 million miles from Sol has been getting noisier in an obvious way for a century. If someone is listening, we have been detected from a very, very long way off.

In this case I'd bet that probes can learn a lot about us without being detected. It's possible to imagine passive flybys attached to/disguised as comets, and passive observation stations on Near Earth Objects. Is getting into position without being spotted difficult? Of course, but assuming a civilization with much higher tech than ours I'd guess it was possible up to about 1970 or so. The probe's builders are willing to wait 100 years to get from one star to another, thus they are willing to take 10-20 years to rendezvous with a comet or NEO so we can't spot them.

A probe which detects intelligent life on Earth from 20 years out has very good options to avoid being spotted. It might opt to simply phone home for instructions and avoid us altogether. This is how I'd program my probe.

Case II.) The planet is inhabited by smart critters, but they have not yet developed useful technology. If they're really primitive, they're impossible to detect until the probe is in-system. Thus it behooves any probe approaching any system to behave cautiously.

The probe drops passive cameras while a light-year or more out and those cameras maneuver in such as way as to avoid pointing a flare at the target solar system. Ideally, the probe wants them to pass through that solar system so that they can photograph all the planets in that solar system (we've already detected those planets because we're a mighty, high-tech, star traveling civilization, right?) Once the passive sensors report back, the probe decides that this planet is still building stone monuments with manual labor, or is even less high-tech. Hiding from this civilization is very, very easy, and since we have at least a thousand years to wait for them to build high technology, we can monitor them very efficiently and remove the monitoring systems without the natives noticing until this planet reaches the technological equivalent of Earth's 1940s. We can then monitor from the previously placed passive systems on nearby planets and moons. Note that Earth has been a Case II planet for most of the past 200,000 years. (Whether a probe would designate pre-Neanderthal humans as intelligent is an open question.)

Case III.) The solar system is inhabited by a high-tech low-energy society. This is difficult to detect from a long way off, and can easily be mistaken for a Case II planet. They might simply be more advanced than the petroleum-burners (poor primitives) in Case I, or they may have taken another development path. They might even be post-technological, (scary) and/or oriented towards nano-tech (really scary.)

If the probe knows, from a long way off, that a planet is Case III, it doesn't go there at all because they will spot it fairly easily. It phones home, hides as quickly/carefully as possible, and waits for instructions. If the probe discovers a Case III society while its passive cameras are in-system, it also hides, and it does not recover the passive cameras. (They stop transmitting and attempt to quietly self-destruct.) If the probe is sure it has been discovered, it accelerates as quickly as possible in a direction that can't lead back to its origin, phones home if it can do so undetected, and (possibly) self-destructs.

The reason for this is that a Case III society can probably become a Case IV society very quickly if they feel like it, but they've chosen not to so the probe-builders don't understand them. Are they timid? Paranoid? Not oriented towards exploration? What's up with them? If they're post-technological God only knows what they've got in their archives. Handle with extreme care until the probe-builders are certain about them.

Case IV.) This is the really interesting society. High-tech and high-energy; essentially equals to the probe-builders. (Probe-builders are Case IV societies.) They've probably got the resources to accelerate a missile to relativistic velocity, just like the probe-builders, and that makes them really dangerous. But they're like the probe-builders and possibly probe-builders themselves, so they're also really interesting. Peaceful probe-builders might want to talk/trade with them.

Fortunately, such societies can be detected from a long, long way away. If our probe detects one, it phones home and hides. We interdict this area of space from our other probes, then we can build appropriate defenses before we send our probe to say hello.

To sum up, any probe nearing our solar system between 1930 and 1970 knows well before it arrives that we are intelligent and that detection is possible, though not probable. It will stay light-weeks or months away, use much higher tech than ours to hide, and will use highly stealthed, passive sensors, which we are not looking for, to learn about us. If the probe arrives anytime since 1970, it knows we will certainly detect any high-tech activity in our solar system. It might hide out someplace a light-month or so away and do some careful probing; it might phone home, and it might simply coast until it is out of detection range and go on to the next destination.

The two real arguments that state we will not find a probe have nothing to do with technology and engineering. They have to do with behavior (there is nothing to lose by being paranoid about meeting other species) and timescales. One would have to be wildly optimistic about the Drake Equation and expect behavior from high-tech aliens that borders on the stupid to imagine that our solar system sees a probe at more than 100,000 year intervals.

Unless we accept Karl Schroeder's idea that advanced civilizations become indistinguishable from nature.

But I'm not buying that argument - it assumes a common pattern of development. Only one has to be a profligate Kardeshev type II or III civilization and we should see some evidence of their activity - unless we are seeing it and just don't recognize it.

flush of chemicals (especially industrial polymers) which can't be degraded yet (if ever).

There is some evidence that some bacteria are able to consume some plastics, like nylon and polythene. However, if they can't it makes for an interesting possibility that these plastics might just be an unequivocal signature of a TUATD that had long left the scene. A probe sampling an alien ocean and detecting these molecules - that would be a very interesting finding.

Only if we ignore hydro, wind and solar power...

I was particularly intrigued by the takedown of Landis' percolation paper, as it was a model that was starting to have the right feel to it- the paradox is all about the signs of intelligence propagating very much like a gas, but a quick glance around the biosphere shows self replicating systems nevertheless have limited ranges, and a glance around the sky naturally suggests that the situation hoping from star to star may be more node-and-edge than petri dish. However, as Wiley suggested, maybe not.

I still have some issues with self replicating probes. I never liked Sagan's version of sitting around fearful of cancerous berserker probes, so I appreciate the dismantling. However, even if they are well behaved, I'm not sure who the user is. If we posit that the probes are something that a civilization does, rather than the civilization itself, which seems like a reasonable variant, and the bulk of civilization expands at a slower rate- and staying home is of course a slower rate too- then said civilization will be in a perpetual state of getting more and more data about less and less directly relevant locales across timescales vastly separated from any entity that initiated the project. Correctly noting that a population of SRPs could visit every star in fifteen million years or the like does nothing to resolve the question of what institution or entity will be sufficiently patient to wait fifteen million years for reports of something interesting that it can then engage in a conversation with 200,000 year end to end lag.

Telescopes have come up in passing, but I think they deserve a closer look. Right now, the apertures of telescopes double in size every 28 years, which is certainly faster and more consistent than any improvements in spacecraft propulsion, and unlike the maximum performance of a probe, which starts to bonk its head on relativity and mass fractions, I don't know of any theoretical reasons why one can't construct truly massive imagers, with truly startling, Google-Earth-from-Alpha Centauri type results, especially as optical interferometry is beginning to come online- and they don't make you wait around for results. It seems to me, again at first glance, that the availability of that sort of insturmentation, combined with the aforementioned lag issues with probes, fundamentally changes the exploratory paradigm- robot emissaries are now things you send to interesting places to gather specific kinds of data, it may or may not be worth it to drag or push along replication machinery, and our expectations of their frequency and the inevitability of our detection of them drop accordingly.

And, again, as others have said, the reports of their observability may be greatly exaggerated. It's been well discussed that all that radio SETI has eliminated is that it is common practice to leave very repetitive narrow band omnidirectional microwave emitters burning all through the galactic night- and we have cause to believe that the accidental transmissions will more and more resemble noise, and/or be more and more targeted. When it comes to probes, it's taken as a given that we'd spot them- but again, all we can say for sure is that there haven't been monstrous numbers of fusion exhaust plumes in the couple decades when we'd notice, and we are still finding cold objects the size of continents in the Kuiper belt- much less some nigh-magical interferometry footballs. We don't have the capacity as yet to tell if a planet is green, much less if it has nightlights (which we want to get rid of) pollution (ditto,) or to determine if it has been terraformed, if the asteroids have been hollowed out, or if some dust cloud is a Dyson swarm or mining tailings or anything else. It's totally within the realm of possibility that each new ratchet in observational power (which is, note again, led by telescopes and not probes)will point out that the galaxy is busy.

Most of those scenarios still suggest a zoo variant, where civilizations that are capable of noticing us don't bust down the door, and that usually gets knocked as requiring uniformity of motive across species, cultures, and timescales. I'm not sure the degree to which that's a dealbreaker, thanks to convergent evolution. As others have pointed out, there are mighty good reasons for civilizations to be polite- not getting relativistically carpet bombed being a big one- and that if, on Earth, we can see emerging consensus about practices regarding being nice to each other when we are still mucking around in orbital space, it doesn't seem unreasonable that such a consensus that includes a degree of discretion could emerge in longer time scales.

I always wonder about the granularity of our observations and their presence, too. Once again, there are places just on the Earth that are ostensibly inside the scope of human affairs where one could camp for decades and not be bothered. Imagine if the Uranians polished their first telescope mirror the day after Voyager flew by and had concluding in the next three decades the solar system was empty. Once again, it's assumed that the scales involved iron out the kinks, but I wonder if anyone has a good way of modelling the sizes of the voids in even ubiquitous interstellar traffic to see if they can comfortable accommodate the phase space of our observations to date.

Karl's re-wilding concept is interesting, but to accomplish activities beyond passive collection means utilizing entropy differences for powering your activities. Even if you are incredibly more efficient you will run up against limits such as the 'Carnot Cycle':

http://en.wikipedia.org/wiki/Carnot_cycle

Those limits will either limit your behaviors or force you into more 'active' means of power generation (and therefore much more visible). Your best bet would be to hide in naturally occurring relatively high energy area's such as:

http://www.space.com/12043-saturn-moon-enceladus-salty-sea-underground.html

of course such energetic moons would be highly attractive to our probes... any sensors you have will need to be hidden as naturally occurring phenomena. If we discover you the mere fact that you are hiding will engender a level of distrust.

A silent, passively observing fly-by probe does provide your best bet. I wonder any calculations have been made with regards to the pluto probe's visibility (disregarding its broadcasts, just its thermal signature) from earth.

New Horizons: http://www.nasa.gov/mission_pages/newhorizons/main/index.html

If our hypothetical probe can reproduce itself, just what sort of stuff is it using to make the copies? Are we talking about some sort of unpacking autofac that docks with a metal-heavy asteroid maybe 10 A.U.'s from the Sun?

Or are we talking about something made out cometary ice? Materials readily scavengeable from the Oort cloud?

You've also got to be careful about your assumptions regarding transmission speed as well. Probes shooting through a system at 10% or even 1% c strike me as being a little . . . fast for this sort of thing. I'd guess something more on the order of 0.1% c or 0.01% c or even slower. Bear in mind that it doesn't matter how fast your probe travels if replication times are an order of magnitude more than the time spent between target stars.

How about something like this: because of mass and energy constraints it's better to send out problets rather than probes. Something on the order of milligrams rather than kilograms. But since you can't pack all the functions you want into something that small, you broadcast several dozen or several hundred different varieties. Once they reach their destination they embed themselves into cometary material and start reproducing. If you have enough different kinds of problets in one comet to reconstitute your probe, fine. If not, once a critical number is reached your problets start shooting copies of themselves into the surrounding space in the hopes of eventually contacting other comets that do have the missing types. Eventually, statistically speaking, they'll succeed.

This may not sound particularly fast, but since we're talking long-term anyway, it doesn't really matter. Also, while the lead times might be long, eventually you'd get to the point where you'd be logging several or several dozen fresh contacts per year every year.

That's not perfectly accurate- coal is known from essentially all geological eras up to the present. There's little bits of Precambrian coal, whose source carbon had to be aquatic and lignin-free, and there are numerous coal beds from the Cenozoic that while presently not commercialized (thanks to their "youth") are still common, and in the US, Cretaceous coal beds are actually far more plentiful than those from the Carboniferous 140 million years earlier- and lignin was certainly digestible by then, relatively speaking. Lignin still isn't very digestible- bit of a niche role- so it's generally suspected that the Carboniferous coal "pulse" had to do with plants bearing more lignin-rich than today to fend off apex insect predators- again, an unlikely condition, but you still make plenty of coal in other conditions. Anytime you have sufficiently anoxic water conditions and plentiful vegetation will suffice.

Isn't it fairly surprising that Earth has been home to complex multicellular animals for around 500 million years and in all that time we are the only tool-using, adaptive-technology developing (TUATD) creatures to have evolved

Not necessarily. The problem is that it's difficult to know how much of the time delay is because the emergence of 'TUATD' is an intrinsically rare event, such that a lot of time usually passes before it happens, and how much is because the emergence is a relatively likely event that simply requires a lot of previous development before its appearance become possible.

Steel-and-concrete skyscrapers have only been around for the last 100 years or so of human existence. Does it therefore make sense to argue that skyscrapers had only a 100/100,000 chance of occurring? And that this chance was the same all through human history (e.g., the probability of skyscrapers appearing circa 20,000 BC was exactly the same as for 1900 AD)?

Okay, if not in black holes, what's the best way of getting really hot, dense matter with efficiency and for a very long period of time- I'd like to think at least a few orders of magnitude past average stellar lifetimes.

Hydro I'll grant you -- within limits. Wind ... high efficiency wind turbines appear to require fairly advanced engineering; we're not talking mediaeval windmills here. Solar thermal is maintenance-heavy and problematic for a variety of reasons; photovoltaic is, again, advanced engineering. The cheapest form of solar, in some respects, is plantations of something that photosynthesizes, grows fast, and can be burned.

Isn't it fairly surprising that Earth has been home to complex multicellular animals for around 500 million years and *in all that time* we are the only tool-using, adaptive-technology developing (TUATD) creatures to have evolved[1].

Your figure of 500 million years is relevant only if there was a possibility for these types of creatures to exist since that time. And while I can believe in intelligent tool-using dinosaurs, I find myself just a bit incredulous when it come to intelligent tool-using dragonflies :-)

Throw in the propositions that the sort of beings you posit can only exist around Population I stars, that these stars are relatively young, and that reasonable trip times for SRP's are on the order of 10,000 years per light-year and you get a scenario where even if intelligent tool-using life is common it just hasn't had enough time to make itself known here out in the boonies yet.

And by then you've forgotten what the question for the ultimate answer is anyway.

Hypercomputation might be possible within a Black Hole of a particular type

http://books.google.co.uk/books?id=3tejKrkWkD8C&lpg=PA146&ots=CMptYHWtBh&dq=black%20hole%20hypercomputation&pg=PA146#v=onepage&q&f=false

"Not really; you're taking fossil fuel deposition for granted"

Sorry, I was specifically wondering why no other technological civilizations had developed on planet Earth between 500 MYA and now.

We've had 500 million years of multi-cellular organisms and in all that time the only technological civilization we know of is us.

I was pointing out that the continued presence of coal and oil deposits (laid down between 300 and 360 MYA) was evidence that no Earth-based technological civ had developed between when those deposits were laid down and the present day.

Given that I believe that any tool-using Utahraptors or whoever would eventually have stumbled across the super-duper energy source that is coal/oil/gas etc, the continued existence of those deposits is evidence that technological civilizations are rare even if multicellular (tool-using?) animals are not.

"In fact, if in general tool-using adaptive-technology developing species have to leap straight from the equivalent of wood-burning tech to nuclear fission, that would be one candidate for the Great Filter: we're well past it, but we're very unusual."

Right. So combine rarity of easily exploitable fuel sources (e.g. fossil fuels) and rarity of technological civilizations and you have your Great Filter.

Another case solved.

Another case solved.

Agreed. This myth is thoroughly busted.

"rarity of technological civilizations"

In fact, ISTR that homo sapiens and our antecedents have been around for around a million years or so. Homo sapiens sapiens you could have fertile young with have been around for (what?) 100 000 - 500 000 years.

But we only developed agriculture and city living in the last 10 000 years and space travel in the last 100 years.

So it took "us" ages and ages and ages of being really primitive before we made the sudden phase transition to "maybe our machines we'll conquer the galaxy".

And it took animals on Earth ages and ages and ages of being just animals before one of them became "us".

I'm now seriously thinking that a combination of "technological civilizations being rare" and "scarcity of super-charging easily-available energy sources" might explain it. We could be the first.

My understanding is that fungi capable of digesting lignin didn't evolve until the Mesozoic, in the Triassic or Jurassic. Fungi do a lot of wood degradation now, as do the termites that also evolved in the Mesozoic.

The issues with lignin make an interesting case study for what will happen to polymers and other industrial chemicals. Unless we deliberately set out to break these chemicals down, it may take quite a while for organisms to evolve the pathways themselves.

I very much doubt we're The First, but we could be early going where intelligent races are concerned. Since the more complex elements are created in the explosions of supernovae, we need multiple generations of star formation to create something we'd recognize as "life." This implies that while the universe is 14 billion years old, life is much younger than that. I doubt there are any intelligent races with histories that stretch back more than a billion years. It's likely that we are the first in this general area of space.

"simply requires a lot of previous development before its appearance become possible"

This is the thing. Evolution isn't progressive, it hasn't been "building towards humanity". We've had the potential for things like us to evolve for several hundred million years. A dinosaur isn't an obviously less complex animal than (say) a turkey. And if it is, what is the magic formula that makes the difference?

"Steel-and-concrete skyscrapers have only been around for the last 100 years or so of human existence. Does it therefore make sense to argue that skyscrapers had only a 100/100,000 chance of occurring?"

Yes, pretty much. Given that if you observed an Earth with hunter-gatherer homo-sapiens wandering around you'd have to wait a long time before they got their act together and built skyscrapers and all the necessary ancillary technology. It could be our lot are particularly thick, or it could be the confluence of factors necessary to stimulate sky-scraper manufacture is actually really quite rare.

So to summarise: my claim is that "given that it took so long for technological civilization to occur on Earth, despite there being multicellular live around for a long time, and given we observe no Dyson spheres, SRPs, etc, it could just be that technological civilizations are incredibly rare."

That's true- but Charlie was saying that coal was a byproduct of lignin evolution going unchecked by lignin digestion, and while there's no doubt that helps, it's not necessary true- we have plentiful recent coals that postdate said fungi, and peat beds turning into coal, and coals older than lignin, and through it all the common denominator is the presence of conditions too acidic and anoxic to allow rot, and that has more to do with hydrology than it does with the particulars of the plant biochemistry. It doesn't seem unreasonable that extraterrestrial plants would evolve a structural polymer that had a nominal resistance to bacteria, simply as an immune feature, and that they'd have wetlands.

Agreed. I was just commenting on the particular issue of why there's so much Carboniferous-era coal.

Is getting into position without being spotted difficult? Of course, but assuming a civilization with much higher tech than ours I'd guess it was possible up to about 1970 or so.

I suspect it's still pretty easy even now. We just aren't that observant, and space is big. Considering the Spaceguard efforts to date, there could be quite a few bus-sized objects even within the inner solar system and/or the belt and we wouldn't know. Probably even manoeuvring, especially if they use something with relatively cold exhaust like an ion drive.

They'd have to be very unlucky to be detected even once, and we wouldn't make much of a single detection (see also: Wow! signal). Multiple detections would probably just get them on the list of minor planets (several times, if they change orbits). It would take a lot of luck and/or activity for us to notice anything.

I think it depends on what you mean by "easy." Would we notice an SRP making an orbital insertion burn near Neptune with a fusion drive? Almost certainly. Would we notice a bus-sized asteroid that was a little hotter than it should be? Almost certainly not.

The problem for a probe is that they must take the most pessimistic possible view of a world's technology. As Vernor Vinge tells us, "...quantum torsion antennas can be built from silver and cobalt steel arrays, if the geometry is correct." In other words, there are probably simple hacks on 1970s technology that would allow a planet to do some really cool detecting "if they knew how" that any given alien race might or might not know about.

Keith, thanks for your kind words. I sounded probably harsher than I meant, and this is a fascinating paper. I still think that SRP's cannot be made perfectly safe from copying errors, and really that was my only gripe. There simply is no way to prevent a mutation on SRPs on a sufficiently long time frame, and then the cat's out of the bag.

But, more importantly, I don't think this issue matters too much (and I should have been clearer about this in my original comment, sorry), because one would have to say that at least one civilization would surely be stupid/brave enough to take the SRP risk, no? And all it takes is one. So I think the rest of your analysis is still fine.

tl;dr: fault tolerance does not invalidate Sagan's analysis, but hubris on a galactic scale does.

If/when the SRPs mutate...then you just redefine "civilization" to mean "That civilization of self-propagating SRPs".
Finding those would satisfy curiosity about extraterrestrial life just as well as finding evidence of their original civilization. And it still raises the question of where the damn things are.

I still think that SRP's cannot be made perfectly safe from copying errors, and really that was my only gripe.

This is one of the reasons any civilization using this strategy of exploration will do very careful mission planning to minimize the number of SRPs they create. (Other reasons to minimize the number of SRPs created include resource usage and not annoying the neighbors.) One of the reasons for periodic stops in SRP creation (like every hundred light years) is to check every SRP for "genetic drift" and correct that as necessary. Overcoming the safeguards against becoming an independent agent or multiplying infinitely probably requires more than one mutation, thus periodic stops to make sure that "genetic drift" has not occurred are necessary.

Maybe we are the SRP's?

I'm feeling fondly reminded at reading "Code of the Lifemaker" from James P. Hogan. Sorry, just had to say that, though it is not productive towards the discussion, except maybe for speed readers, because it describes a scenario of a SRP crashing on Titan due to some malfunction aeons ago, whose descendants chaotically evolved into a society resembling the middle ages when an exploration team from our times arrives there. Funny stuff. Really :)

I'll repeat my comments from Karl Schroeder's page:

1) With respect to SETI, given the energy requirement and the time and distance involved, there's a very good chance that virtually all interstellar communication will by highly compressed. I have it on very good authority that highly compressed signals would look like noise, absent the algorithms to decode them. So, we may receiving them already. Again, given the constraints noted above, such signals would be fairly infrequent so they might not manifest a repetitive pattern.

2) Peter Watts' Blindsight demonstrates the possibility (noted by neurologists/neurophilosophers like Damasio) of non-sentient intelligence. I leave it to readers to identify the implications of such intelligences when it comes to civilizations communicating.

3) Energy production. One of the major issues with renewable, wind and solar in particular, is storage. Hardly a week goes by when I don't find a new article about major progress in storage technology. A quick search of Physorg (a science news aggregator) shows pages of stories relating to new technology in capacitors and batteries.

The beat goes on. None of us will know until real contact is made.

The problem with that is, how do you handle SRPs learning from experience?

An instance of the problem: Florence is an A.I. and so has the Three Laws built in.

Alex R @ 107: Since the more complex elements are created in the explosions of supernovae, we need multiple generations of star formation to create something we'd recognize as "life." This implies that while the universe is 14 billion years old, life is much younger than that. I doubt there are any intelligent races with histories that stretch back more than a billion years. It's likely that we are the first in this general area of space.

Actually, you can get fairly "metal-rich" stars (i.e., with high abundances of elements heavier than H and He) early on in Galactic history. We know (from observations of very distant quasars, among other things) that you can get solar levels of complex elements in the centers of large galaxies within the first couple of billion years. Even in our part of the galaxy, it's possible to find stars as metal-rich as the Sun but several billion years older. (These are more common towards the center of the Milky Way.)

The trick is that the time from the formation of massive stars to the first supernova is a few million years. So if you can keep star formation going at a high rate in the same region, without diluting it with too much "pristine" gas from outside, you can build up a lot of heavier elements in only a few hundred million years.

On the other hand, not having frequent supernovae in the vicinity is also probably a prerequisite for life as we know it.

"... photovoltaic is, again, advanced engineering."

It doesn't have to be all that advanced: How to Build Your Own Solar Cell.

The causes of the industrial revolution - for that is the Great Filter we seem to end up with - are hotly debated. At one end of the spectrum are single-cause advocates like Oded Galor, and at the other are the "long chain of unlikely coincidences" people, such as Pomeranz and Wong. (Property rights had to be insecure - but not too insecure, labour had to be scarce - but not too scarce, transport had to be easy - but not too easy, a vast new set of resources had to turn up - but not too soon or too late, and so on.)

Anyway, we're not through that filter yet. We can't even boil water with a fusion reactor, or create a self-replicating (physical) machine, let alone create one that will work in interstellar space for 1,000 years at a time. And we're starting to dick around with DNA in big ways without much idea of what could happen. We could end up in a story by Vonnegut.

On the thread about blackbody radiation ruining stealth, e.g. @Charlie Stross (#77):

"Running nuclear power plants in the outer solar system are going to stick out like a sore thumb thanks to their infrared emissions."

You can shield blackbody radiation emitted in specific angles (such as the direction of the planet you're hiding from). For example, you can put a thermal insulator on one side of you probe, and not on the other. Then most of the heat is dissipated on the unshielded side; the surface on the far side of the insulation will stay cold and radiate far less.

A back-of-the-envelope example. Say you're trying to dissipate heat from a 50 kW/m^2 plate source. Say one side is a conducting blackbody plate (A), and the other is a 1 meter thick layer of insulating silica aerogel (optimistically 0.004 W/m*K [1]), sandwiched by another blackbody plate on the far side (B). Then at steady state

T(A) ~= 970 K (realistic) T(B) ~= 89 K P(A) ~= 50 kW/m^2 P(B) ~= 3.6 W/m^2

[1] http://en.wikipedia.org/wiki/Thermal_conductivity#Experimental_values

The "far" side is ten times hotter than the "near" (stealth) side, so emits four orders of magnitude more blackbody radiation. (T^4 in Stefan-Boltzmann's law)

You can easily get two more orders of magnitude by playing with radiator materials -- make the unstealthy side very black, and the stealthy side very light-colored/shiny:

http://www.engineeringtoolbox.com/emissivity-coefficients-d_447.html

So there's 6 orders of magnitude reduction in blackbody emissions (in one direction, that is over 1/2 of a solid sphere), using only 20th century tech and pretty modest parameters.

Background: http://en.wikipedia.org/wiki/Stefan%E2%80%93Boltzmann_law

?

This is the thing. Evolution isn't progressive, it hasn't been "building towards humanity". We've had the potential for things like us to evolve for several hundred million years. A dinosaur isn't an obviously less complex animal than (say) a turkey. And if it is, what is the magic formula that makes the difference?

My point was that technologically capable intelligence may only become possible after a minimum amount of development, not that it becomes inevitable.

Actually, you're partway towards agreeing with me: you start off by assuming that "complex multicellular life" is, somehow, a necessary prerequisite, and thus that it's only the last 500 million years that count -- as opposed to the argument made by some that it's been at least 3 billion years since life first appeared, and intelligent life could have arisen at any time during that period, and therefore sapience must be really, really rare.

But even accepting that it would be possible for trilobites to give rise to something like us (in the sense of sapience, not body plan!) given, say ten million years, are you really arguing that the potential was the same as it was for mammals or birds during the last fifty million years?

A more plausible model is an increasing probability (or "potential") with time, simply because that allows for the possibility that sapience requires several prerequisites (e.g., multiple sophisticated senses, something like a basic nervous system, complex social life) to accumulate before it becomes possible. This could still mean that the late-time probability is very, very small, of course; I'm really only questioning the "it could have happened (with equal probability) any time in the last 500 million years" argument.

(As for dinosaurs and turkeys -- my impression is that dinosaurs were, on average, probably less intelligent than birds -- e.g., lower brain/body mass ratio, though that's a terribly crude measure, and it's not clear how many dinosaurs had brains filling their skulls the way mammal and avian brains do. There were some dinosaurs with relatively large brains; curiously, these appear rather late in history. If you want to find dinosaurs that were probably as smart as birds, you're better off looking in the Cretaceous, not the Triassic.)

Hydro I'll grant you -- within limits. Wind ... high efficiency wind turbines appear to require fairly advanced engineering; we're not talking mediaeval windmills here. Solar thermal is maintenance-heavy and problematic for a variety of reasons; photovoltaic is, again, advanced engineering. The cheapest form of solar, in some respects, is plantations of something that photosynthesizes, grows fast, and can be burned.

Also, don't forget tidal, geothermal and peat.

In other words, it is not convincing that the only way to nuclear energy is through fossil fuels extravaganza. Do we really need 7 billion people to get to fusion?

How about the burning of biomass along with the creation of charcoal? With a good enough plantation program it might be enough to bootstrap an industrial revolution towards the harvesting of natural gas, biofuel or fission.

Someone needs to write a book about industrial revolution with no coal.

winkwink

Exactly. Yet another reason why SRP builders, if they exist, will be very, very careful.

(Once again, I don't see the need to build SRPs. I'd expect single use probes going to specific places chosen by telescope.)

BTW, Freefall is one of the best webcomics out there, and IMHO deserves some kind of award.

I have several problems with that paper.

First, if the probability of "probes being here" is that high, then I see no reason to assume that they are not here. So how would we recognize them? Are sharks something that arrived in probes?

Second, if the probabilities are that high, then how do multiple probes interact when they encounter each other? What does that do to further replications? (War sucks. So from this I conclude that war sucks for self replicating probes.)

Put differently: any sufficiently advanced technology is orders of magnitude simpler than "ordinary" biology.

No. The efficiencies of photosynthesis and biomass production are too low. Recall that using quality biomass - wood - effectively caused the deforestation of Britain. Conversely, Japan, a country that maintained forests for building materials [J. Diamond Collapse] did not effectively industrialize during this period.

We need fossil fuels until we can transition away from them to other ones. The green fantasy is that our technology will become efficient enough to work with biomass production (and still allow us enough food to eat well too).

What is there to be wrong about in biology?

In fact, we don't have much of a "theory of biology" beyond the Hardy-Weinberg principle. Everything else is bits and pieces that are ad hoc.

Evolution itself isn't a scientific theory, but a proto-scientific theory -- a collection of facts and reasonable explanations that haven't been quantified (beyond Hardy-Weinberg). Biology is still where physics was with Galileo. Biology's Newton is a shade from the future.

So, I'd put it this way -- about what biology knows, it's unlikely to be wrong. But we may find that a lot of what we know is about phlogiston, when what we really need to be studying is oxygen reactions.

Thanks for commenting on my paper. While I appreciate your points, I don't see how they are arguments against the paper. You seem to be offering arguments in support of the paper's undercurrent: that SRPs and consequently ETI are probably rare.

I suspect there's an "arrow of increasing intelligence" in evolution. Intelligence (massive information processing to survive in a complex environment) appears to be expensive both in terms of energy and nutrients. It also give some competitive advantages in certain environments (on land, where resources are really good. Less, say, in the abyssal ocean).

The "arrow" is due to the fact that intelligent consumers are dangerous. They strongly direct the evolution of the animals and plants they consume, as well as the herbivores and predators with which they compete for resources. This pressure can favor increasing intelligence among the other species (and other sophisticated responses in general, such as more interesting plant defensive compounds, defensive and offensive symbioses, etc). Over time, animals (and possibly plants) get more sophisticated, simply through interacting with more intelligent species, and through investing more and more resources into information processing.

However, this is a very, very gradual process. There are lots of ways to be stupid and survive very well indeed. The fact that our biosphere is still run by prokaryotes (as it has been for four billion years) is strong evidence of that. More generally, intelligence is expensive, and it can only occur where there are resources to support it. In low resource situations, KISS is a better alternative.

The only way one could get a hypothetical intelligent trilobite is for something else to show up in the system (an Elder Thing, for example) and provide strong selection pressure for smarter trilobites.

I suspect there's an "arrow of increasing intelligence" in evolution.

Define intelligence. Then we can try to measure it today and in the past.

It's even simpler than that. On the side facing the planet you want to shield from, you have a disc with circulating liquid helium cooling its surface. A few years ago I did a series of calculations for sci.military.moderated on stealthing ballistic missile warheads using liquid nitrogen sheathing. It's a tech that is being actively pursued right now.

"Cheap FTL travel could provide a similar means of making galactic colonisation impossible."

????

If FTL looks feasible then the Fermi Paradox becomes critically serious. We all await the neutrino speed results...

Certainly: intelligence is processing power, including things such as numbers of separate reflexes, separate instincts, or use of learning as part of an adaptation to the environment.

If you think only humans are intelligent and everything else is dumb, you're stuck back around 1940. When talking about an arrow of intelligence, I'm pointing to is the general increase in things like encephalization quotients through time (size of brain to size of body).

For example, birds are theropod dinosaurs, but most birds today are substantially more intelligent than all but the smartest theropods. Similarly, the theropods of the Maastrichtian Cretaceous appear to have had bigger brains (proportionally) than the first theropods in the Triassic. Even Carl Sagan noticed this.

This affects plants as well: there is probably greater chemical diversity, greater symbiotic diversity, and probably greater signalling diversity now than there was in the Cretaceous (so far as we know), and certainly more than there was in the Triassic.

To reiterate the point, intelligence is often useful when your competitors, predators, and prey are also intelligent. Since there's a high metabolic cost to intelligence, I don't expect it to emerge out of nowhere, any more than I'd think that anti-spam programs were necessary for a 1980s PC. If you know about the Red Queen theory in biology (the idea that coevolving predators and prey have to keep evolving just to stay still), you could say that intelligence is the product of Red Queen-style evolution.

The WAP + multiverse hypothesis is the one I keep coming back to. Just as it explains (for peculiar values of "explain") the fine tuning problem, it handily takes care of the Fermi Paradox.

Downside with that is the 'facing the planet you want to shield from'. You still have to dump your waste heat somewhere...

How many of our space probes have taken 'looking back at earth' famous 'blue dot' photo's? Its become fairly routine to have our probes look back, both for the posterity shot and recently to calculate what the earth looks like to remote observers at both visible light and infrared spectrum.

New Horizons did it from deep space, as did the voyagers (from really deep space), and the 'Deep Space 1' probe. An active SRP is going to have be further and further out to escape detection. From the earth I wonder if an infra-red scope would see anomalous cold spots...

Liquid nitrogen sheath might work for quick ballistic missiles but for an active SRP, it will still have to dump the heat from some side leaving a hot spot 'sore thumb' in deep space.

Hydroelectricity seems a perfectly fine way to develop/maintain industrial civilization if the population:hydro resources ratio stays low enough. The water wheel is more than 2000 years old. The water turbine is about 180 years old. Hydroelectric power plants began operating in the 1880s. Once you have hydroelectric power you can use it to refine high-purity copper and build more and better hydroelectric plants, up to the limit of the available hydroelectric resources. Hydro is renewable, dispatchable, and cheaper than coal. The environmental drawbacks are real, but none (IMO) worse than getting equivalent energy from fossil fuels. The most serious limitation is that you can't add more once you run out of appropriate geography.

The United States gets nearly 10% of its electricity from hydro. If the US had a population the size of Canada's, it could get almost all of its (current profligate) electricity use from hydro power alone. If the US had never enjoyed a glut of fossil fuels, it would have probably developed to take a great deal more care in economizing energy use in transportation, climate control, and industry, so the electricity could be used to do more in industry that's currently handled by fossil fuels.

Radioactivity and the potential of atomic energy were discovered before WW I. It seems quite likely to me that fission would have eventually been discovered and exploited as an energy source even if the 20th century didn't include a fossil fuel boom, population boom, or total war between leading industrial powers. In fact fission probably would have benefited from being developed slower and later, after health physics was better developed and without the military pressure to build a superweapon ASAP.

Um, let's compare string theory and evolution, shall we? According to modern cosmology, we don't know what 96% of the observable universe is, except that it's this stuff we can't detect (otherwise known as fairy butts, or if you believe Terry Pratchett, the 96% is the paperwork on the 4% we can observe).

Compare this with evolution, which is among the most harshly tested theories on the planet. Evolution has been observed in the lab, in the field, and in the fossil record. It has also been turned into a hypothesis-driven science through cladistics.

Evolution is the fundamental theory of biology, not Castle-Hardy-Weinberg.

Want to compare that with, say, string theory? Or quantum theories of gravity and time?

Active-cooling has been suggested for stealth aircraft to reduce their IR signature which is up there in lights given the capability of modern infrared detection systems employed in close-in antiaircraft multispectral sensor heads. The idea is to use liquid hydrogen to fuel the aircraft but bleed the H2 through a network of pipes on the underside of the fuselage before burning it in the engines. You can't do anything about the jet exhaust except to diffuse it and shield it from observation from below but that's a solved problem generally for stealth aircraft.

Occam's razor is all very well when things are consistent

Or, as intelligence analysts have cause to appreciate, when someone isn't actively trying to use the razor to fool you. Anybody for Manichaeism?

Occam's razor is all very well when things are consistent

Or, as intelligence analysts have cause to appreciate, when someone isn't actively trying to use the razor to fool you. Anybody for Manichaeism?

A bit like the best way of stealthing an aircraft from the ground in daylight is to illuminate its underside

Running nuclear power plants in the outer solar system are going to stick out like a sore thumb thanks to their infrared emissions.

I would disagree if we're talking about hiding the IR from Earth. A big (probably several km to tens of km diameter) shallow cone consisting of multilayer insulation with a high-IR-reflectivity inner layer and a low-IR-emissivity outer layer between the power plant and Earth with the apex on the Earth side should do the trick.

I tried to do an ASCII graphic of this, but it didn't work.

See my comment at #143 regarding deep space probes looking back home...

"Downside with that is the 'facing the planet you want to shield from'. You still have to dump your waste heat somewhere..."

Yes, away from the planet you are trying to hide from. Active liquid helium cooled plate planet facing side, and red hot cooling fins the other.

Benjamin Rosenbaum's "The House Beyond Your Sky", found in "The Ant King: And Other Stories".

Certainly: intelligence is processing power, including things such as numbers of separate reflexes, separate instincts, or use of learning as part of an adaptation to the environment.

Now you give me a list of ill-defined stuff. Give me something we can actually measure, and measure it in the far past as well as today.

If you think only humans are intelligent and everything else is dumb, you're stuck back around 1940.

Where is this coming from?

When talking about an arrow of intelligence, I'm pointing to is the general increase in things like encephalization quotients through time (size of brain to size of body).

Brain-to-body mass ratio can be measured, but you have to prove it is correlated to intelligence (elephant vs. mouse seems to disagree). And for that, you need to define intelligence first.

For example, birds are theropod dinosaurs, but most birds today are substantially more intelligent than all but the smartest theropods.

Do you have living dinosaurs to compare with birds? I'd like to see them...

Similarly, the theropods of the Maastrichtian Cretaceous appear to have had bigger brains (proportionally) than the first theropods in the Triassic.

And this has to do with intelligence what?

Even Carl Sagan noticed this.

Is this argument from authority?

This affects plants as well: there is probably greater chemical diversity, greater symbiotic diversity, and probably greater signalling diversity now than there was in the Cretaceous (so far as we know), and certainly more than there was in the Triassic.

Now you have to prove that diversity has to do anything with intelligence.

To reiterate the point, intelligence is often useful when your competitors, predators, and prey are also intelligent. Since there's a high metabolic cost to intelligence, I don't expect it to emerge out of nowhere, any more than I'd think that anti-spam programs were necessary for a 1980s PC. If you know about the Red Queen theory in biology (the idea that coevolving predators and prey have to keep evolving just to stay still), you could say that intelligence is the product of Red Queen-style evolution.

Great Zombie Jesus! What are you talking about??? I don't know what you mean by intelligence.

"I don't know what you mean by intelligence."

Problem solving ability. And if brain:body mass ratio has been increasing over time that is quite significant

Which does appear to me to be somewhat irrelevant to the issue at hand. As already mentioned more than once, what you want to do is to radiate it in some other direction.

You might want to look at the plans for the James Webb Space Telescope, which is doing almost exactly the opposite - it's trying to shield its intensely cold (40K) optics from Earth and the Sun. The JWST is also putting an insulating structure between itself and the Earth, while allowing itself to radiate in all other directions.

Yes, it's obvious you don't know. I'm sorry to be both rude and blunt, but I not going to continue to argue with aggressive ignorance.

Things like numbers of reflexes and instinctive actions can be counted (see, for example this book for examples of the technique in ants).

For encephalization quotient, see Wikipedia reference. Note that if encephalization quotient doesn't matter, then we have to assume that birds are dumber than dinosaurs, with common ancestry and bigger relative brain sizes. That makes even less sense. And note, I'm not comparing parrot or octopus intelligence to mammalian intelligence, I'm comparing birds to their ancestral relatives. You can see the same pattern if you look at the brains of paleocene mammals vs. modern mammals: the only modern mammals that have such proportionally small brains are (apparently) some marsupials.

Again, my apologies for being rude, but this feels like I'm feeding a troll. There are plenty of references out there if you're genuinely interested, and I'll be happy to talk about it further after you've read them.

"I don't know what you mean by intelligence." Problem solving ability.

Ants can solve problems quite well, for a given set of problems. Are they intelligent?

Slime molds can solve problems very well (again, for a given set of problems). Are they intelligent?

My answer to both is "no." "Problem solving ability" therefore does not define intelligence very well or accurately, I'm afraid to say. (Partially because you're using English, which is a very fuzzy thing to define things in.)

In that case: "General problem solving ability" or even: "Novel problem solving ability"

Take your pick

Neither of which are particularly definitive. (Again, partially due to English: you know what you mean, and you use the English words for it, but they're fuzzy and imprecise.)

I've said this before, but the best workable definition of "intelligence" I've seen is: the ability to add time-separated event sequences to the decision making process. Which is a verbose way of saying "to learn cause and effect." (This is different from rote memorization and simple good-bad neural net manipulations.)

Yes, it's obvious you don't know. I'm sorry to be both rude and blunt, but I not going to continue to argue with aggressive ignorance.

You started it.

Things like numbers of reflexes and instinctive actions can be counted

It's not as easy as you think, but let's assume we can agree on what is an instinct and measure it (remember there is still much we don't know about how brains work). Now tell me how are we going to measure the reflexes and instinctive actions of extinct species. I'm all ears.

then we have to assume that birds are dumber than dinosaurs, with common ancestry and bigger relative brain sizes. That makes even less sense.

ASSUME makes an ASS of U and ME. I don't assume anything about dinosaurs' intelligence, I don't have enough information. It is you who assume there is a rising "arrow" of intelligence.

Again, my apologies for being rude, but this feels like I'm feeding a troll

I feel like I'm feeding one. I mean, it looks like you have both a great definition of intelligence and a way to measure it in dinosaurs. Stop teasing us. Or forget about us, you have Noble prize level material, go publish it.

Cluelessness is not necessarily due to trolling; in this case, you both appear to be talking about different things, or at least from very different angles. Please both consider that when commenting.

In what way is "General problem solving ability" imprecise or ill defined? A creature encounters a problem it has not met before, and it's intelligence is measured by how well it comes up with a solution (or not). The problems range from simple eg shortest route to a bit of food, to a Theory of Everything and beyond.

Define "general problem." Define "solving."

Or not; I don't really care to get into a semantics argument with you. But your terms are imprecise, and simply repeating them in different orders isn't going to change that.

Anatoly, please calm down.

"I've said this before, but the best workable definition of "intelligence" I've seen is: the ability to add time-separated event sequences to the decision making process. Which is a verbose way of saying "to learn cause and effect." (This is different from rote memorization and simple good-bad neural net manipulations.)"

Your definition suffers from exactly the same kind of imprecision.

Sorry, I just get a bit agitated when people start talking about grand concepts like Intelligence as if we really know what they are. And with such confidence to boot! Usually people talk like that about god.

If you know about the Red Queen theory in biology (the idea that coevolving predators and prey have to keep evolving just to stay still), you could say that intelligence is the product of Red Queen-style evolution.

Human intelligence (post the cultural explosion) may be a bit different, but I agree with your general argument. As you state earlier, brains to handle useful information processing are costly. Proof of that is seen with domesticated animals that reduce their brain sizes compared to their wild ancestors, probably due to to teh lack of survival based processing needed.

Saying "general problem solving ability" or "novel problem solving ability" doesn't help.

There are many problems that machines or non-human animals can solve better than humans. But a human can always say that those problems are not novel.

There are many synthetic puzzles you could invent that have never before been seen before, and construct them in a way that makes it much easier for machines to solve them than humans (either because they rely on numerical manipulation or exploit common human cognitive biases). But a human can always say that those problems are not general.

Intelligence, like beauty, is not suited to scalar quantification and its recognition is culturally dependent. Trying to construct some universal intelligence scale that can encompass worms, machines, humans, and aliens seems as misguided as 19th century scientific racists constructing theories to explain the "fact" that European people are the best-looking people on Earth. Why should alien thought resemble human thought any more than the human body resembles algae? "Starfish alien" isn't nearly a strong enough term to suggest an intelligence that shares no cultural or evolutionary history with humanity.

"Saying "general problem solving ability" or "novel problem solving ability" doesn't help.

There are many problems that machines or non-human animals can solve better than humans. But a human can always say that those problems are not novel."

Those problems are not of the "general" category. AFAIK there is no machine that exists that is a general problem solver. Animals have general problem solving ability - to a degree. That degree is a measure of their intelligence.

What would one of these advanced probes look like? What sort of detectable signature would they have?

Well - this is all blue-skying mind you - I'm guessing that the parts that get closest to us aren't all that physically impressive. In fact, it looks like a cloud of dust or maybe there's some specularity that makes it look like ice particles. That's because each physical sub-assembly is maybe a hundredth of a millimeter long by perhaps a thousandth of a millimeter thick and comes equipped with some rudimentary sensory equipment along with some modest processing power and a bit of memory. But combine a few million of these things into a network and all of a sudden you've got some decent computing resources, enough say to achieve resolutions of less than a meter from Earth orbit. Since it's solar powered and solar propelled (if the main probe is going to spend over a thousand years between targets, what's another ten or twenty years to close get to the interesting places in-system with light pressure?), you're not going to see much of an energy signature.

Yeah, the bigger part of the probe, the von Neumann part with the really big brain might look more impressive. But there's no point in it lugging all that extra mass dedicated to reproduction around when it doesn't contribute anything at all to the local missions, not when all you really need is bare sensory capability and some power and communications to send data back.

So color me skeptical in regards to how visible such a probe would be from the ground. Or even from just a few kilometers away.

a lump surrounded by a fine cloud of ice particles..... the comets theyre WATCHING US!!!!

Using the word "probe" to designate what we're talking about is making semantic hash of the conversation. What Keith Wiley is really talking about is not a probe so much as an "aircraft carrier" for probes. Calling it a "self-replicating probe carrier" makes a lot more sense in this context, because the "Self Replicating Probe Carrier" might be a couple miles long, and it would hide a couple light-months out while much smaller devices did the actual probing.

So what we're really looking at is probably three levels of hardware; a big carrier for voyages between the stars, several smaller craft designed to carry the packages from someplace in the Oort cloud to someplace close to Earth's (or Jupiter's, or Mercury's) orbit, and the actual science packages you're talking about, which will be placed on a Near Earth Asteroid, or in orbit around a planet, or on the dark side of the moon...

Thus the part we're most likely to see stays far away from us, while the part we're least likely to see - your cloud of sapient dust - goes into orbit around the earth.

Alex, what is this big monster probe of yours made of anyway? You make it sound like it has an alloy composition similar to Kinnison's "Dauntless". I think that's rather . . . farfetched.

Would we notice an SRP making an orbital insertion burn near Neptune with a fusion drive? Almost certainly.

I'm not so sure. Which of our existing projects would (a) detect this at all, and (b) detect it in a way that would be credibly ascribed to ETI? What fraction of the sky do these projects cover each night? (Either now or at any time since 1970.)

Even with a highly-visible hot fusion drive, I suspect the probabilities of our detecting it at all are low and detecting it in a way to recognise it even lower. We just don't look up that much. Spaceguard is woefully inadequate.

Similarly for the pale blue dot / blue marble photos; there's only a small number of them, taken from only a few directions with narrow fields of view. They're also singletons, where a series would really be needed to show most anomalies.

The problem for a probe is that they must take the most pessimistic possible view of a world's technology. As Vernor Vinge tells us, "...quantum torsion antennas can be built from silver and cobalt steel arrays, if the geometry is correct."

Well, that's really your case III, isn't it, not case I. As you write, difficult to detect from a long way off, so they might take it into account anyway, but a different situation to case I...

I have no idea what the main probe is made of. It just seems obvious to me that the "probe" you use to investigate a planet is very different from the "probe" that travels from star to star. One carries the other.

The big "probe carrier" would probably customize the smaller probes en route as it learned more about the target solar system. I'm assuming that the probe carrier has to be very large to carry enough fuel to get from star to star, though just how big it would be depends on the drive technology.

"Would we notice an SRP making an orbital insertion burn near Neptune with a fusion drive? Almost certainly."

This made me realize that I don't actually know how much monitoring we do, in other words what we would notice.

We would notice this because some astronomer would notice? Do we have big powerful telescopes pointing at each of the planets all the time or can we assume this would be noticed because a backyard astronomer could pick it up? In other words, the limit for what we might notice would be set by our most powerful telescopes, but the limit for what we would certainly notice would be set by less powerful telescopes?

Guys, what might make even more sense is that the carrier keeps on trucking at, say, 0.2c, and drops off a bunch of probes and a telemetry transceiver station that slow and enter the system they're to study.

Have a look at "Slow Train to Arcturus" for an idea of how this might work for an interstellar colony ship.

So what we're really looking at is probably three levels of hardware; a big carrier for voyages between the stars, several smaller craft designed to carry the packages from someplace in the Oort cloud to someplace close to Earth's (or Jupiter's, or Mercury's) orbit, and the actual science packages you're talking about, which will be placed on a Near Earth Asteroid, or in orbit around a planet, or on the dark side of the moon...

I have a different model.

Interstellar travel is energy-intensive so you want to make the payload as small as possible or leave the engine back home and use beamed power (e.g. a microwave or laser sail). So what we're looking for is a small sail-driven capsule containing a seed, which on arrival will decelerate and go to ground in the innermost debris belt of a star system (needs mass, needs sunlight) where it will take its time building a factory for compact, stealthy science packages to fling past the local planetary bodies (again, using solar sails with their reflected light directed out of the plane of the ecliptic to reduce visibility).

If the locals look dangerous, the probe factory attempts to self-destruct immediately. If not, then after establishing that there are no obvious indigenous threats, the probe factory "phones home" to the star system it came from and requests a download of the full blueprint for the detailed exploration kit and interstellar seed manufacturing plant. At no point should this information be available in a star system obviously inhabited by spacegoing strangers. Only once a target system is confirmed as being safe should the probe acquire the information it needs to self-replicate and propagate to the next star system.

Also, if a probe factory fails to phone home and confirm a system is "safe" within a set time after its arrival, the parent factory should self-destruct, just in case something nasty is back-propagating along the exploration tree.

I think we completely over estimate our ability to detect stuff out there, and fail to understand, as Douglas Adams so succinctly put it, Space is Big. We don't even have a properly integrated and funded Earth Crossing Asteroid detection programme.

While a fusion engine plume decelerating something into a solar orbit is probably detectable pretty much anywhere in the solar system, somebody has to actually be looking in that direction to see it.

I suspect a GSV with all the lights on could well cruise in and sit in the Earth-Moon L1 without anyone noticing.

'Active-cooling has been suggested for stealth aircraft to reduce their IR signature'

96% of the universe turning out to be made up of actively stealthed complex baryonic matter would neatly resolve both the dark matter and Fermi Paradox controversies. It's also consistent with observations of the Bullet Cluster, providing you assume that the dark matter was manoeuvring to avoid the high-radiation area, as opposed to following a ballistic trajectory unaffected by it.

Are you suggesting Dark Matter is probes?!?

This is a great idea for a story.

I think I just solved the paradox.

There are no SRPs because it is safer, easier and more efficient to use one solar system as a central factory, build all probes there and send them directly to all stars in the galaxy.

Pros: 1. You don't risk infecting the universe with cancer. 2. You have full control of the factory. 3. Probes don't waste time stopping and replicating, but go directly to their targets.

There are ~600 billion stars in the Milky Way. Let's say we need 600 trillion probes, assuming they are really-really faulty. Let's say each probe weights million tons. (I'm being atrociously generous here, with mass and fault rate).

10e9 kg X 600x10e12 probes = 6x10e23 kg total mass of probes.

Mass of Mars is 6.4185x10e23 kg.

I'm done here.

Do you sell SPAM?

If we're lucky, we might get the Spam, Spam, Spam, Spam, eggs, chips and Spam?

[sidebar] Quite happy to mock the spammers.[/end]

Blimey, I think this spammer went full AI.

I was trying to go Monty Python on his posterior!

Intelligence is the property that a being can display by providing a useful definition of the word "intelligence".

Nah. It just copied some text from upthread. A trick that is surprisingly effective.

Unless it's a probe trying to make contact.

There are no SRPs because it is safer, easier and more efficient to use one solar system as a central factory, build all probes there and send them directly to all stars in the galaxy.

There's a problem with that -- you're assuming that you can build probes with a mean time between failures greater than the the time it takes to directly reach all stars in the galaxy.

If your probes are reliable enough to reach stars c. 150,000 light years away (which at 10% of c would take 1.5 million years) then yes, one solar system will do (as long as you don't mind launching a few hundred billion probes). Otherwise ...

You may do better with a hybrid model: just one star system in a thousand hosts a factory, which probes 1000 nearby systems and curates the report of the results, then identifies an ideal candidate for another factory near the edge of its reliable travel distance and sends a new factory there. In which case only 1/1000 visited star systems gets to host a factory (or 1/100 or 1/1,000,000; it's a trade-off between your engineering reliability and the distance you're surveying).

There's a problem with that -- you're assuming that you can build probes with a mean time between failures greater than the the time it takes to directly reach all stars in the galaxy.

I find it no less plausible than the ability to build SRPs.

However, let's assume we have the SRP technology. Do we want to send it into the Great Unknown where some nasty aliens may lay their hands on it? No! The tech in the probe should be the most primitive stuff that gets the work done. What I'd do is keep the replicators in my Solar system, use them to build huge non-replicating probes with thick shells and lots of redundancy, and send those.

You may do better with a hybrid model: just one star system in a thousand hosts a factory, which probes 1000 nearby systems and curates the report of the results,

In which case that factory will surely NOT be in an inhabited system.

How about:

Increasing intelligence => longer term planning.

Also, if a probe factory fails to phone home and confirm a system is "safe" within a set time after its arrival, the parent factory should self-destruct, just in case something nasty is back-propagating along the exploration tree.

I would suggest that a failed probe highlight a system for further scrutiny. Given the number of probes we've lost to natural causes (lots) versus those lost to enemy action (none that we know of), I'd expect quite a few long-range probes will be lost to ordinary hazards such as mechanical failure and radiation damage. On the other hand, if a system eats a dozen probes and none of them even radio back first, that looks like a problem.

Since our hypothetical enemies by now almost certainly know where the nearest replicator factory is, simply by backtracking the probes, it may be more useful for the factories farther up the line to go silent. The factory which has already poked the suspicious system might turn itself into a trap and wait a few thousand years to see what shows up to poke back.

Having said that, I too like the few hub factories and many probes model. Not only does it seem efficient, but many astronomically interesting neighborhoods aren't very well suited for factories.

"Would we notice an SRP making an orbital insertion burn near Neptune with a fusion drive? Almost certainly."

This made me realize that I don't actually know how much monitoring we do, in other words what we would notice.

We would notice this because some astronomer would notice? Do we have big powerful telescopes pointing at each of the planets all the time or can we assume this would be noticed because a backyard astronomer could pick it up?

In other words, the limit for what we might notice would be set by our most powerful telescopes, but the limit for what we would certainly notice would be set by less powerful telescopes?

I'm pretty sure most big telescope time is dedicated to looking at things other than solar system objects. So my money is on the backyard setups: they're numerous, flexible, and quite capable of detecting your average blazing multiterawatt drive putting a ship into orbit around Neptune. If, of course, the drive isn't blazing out those terawatts as visible light, or if the light is highly directional and not pointed at us (that might be a good thing), it wouldn't be seen.

I just don't buy the whole "stealth" paranoia of this whole SPR debate. It all sort of assumes a very Earth-centric viewpoint - meaning, what could be detected by a comparable to 21st century Earth technology, and then imagining a worst-case, North Korean-type reaction ("we will find you and kill you, whoever the fuck you are who sent this probe!").

First of all, 99.9999999(or so)% of star systems encountered will be 'inert'. Blowing all sorts of energy on unnecessary countermeasures seems silly.

If a probe encounters the extremely rare system that's special (sings of life), well, acting all "stealthy" will still likely be unnecessary, or, if the system is well and truly unique (meaning, inhabited by not just life, and not just intelligent life, but by really really intelligent life, maybe way more intelligent than your probe, who might get all kinds of suspicious with it's pathetic attempts at stealth, etc.), might be downright counterproductive.

Presuming that a society that could send out such SPRs could make them pretty damn competent and independent (say, with a full suite of "first contact" variables and such already programed in, among a whole shitload of other stuff - I mean, these are super smart ETIs, right?), you would figure just a modicum of subtlety would be all that was necessary.

Meaning, don't show up in a new system all full of "fire and glory", the more to freak out any primitive intelligences that might be there (on the rare chance that's possible and that your society has some sort of "prime directive"). But otherwise, there would be no point in acting particularly elusive from any locals that may be capable of discerning your presence.

Jeez, you'd think, in fact, that the major point of your entire SPR mission would be to find and stir things up with just such locals!

I mean, I'm sure any self respecting SPR would take minimal precautions, just in case it was discovered by galactic psychotics, and in that very rare and almost unbelievable case, my have a self-destruct mechanism ready to go.

But otherwise, you would think the SPR would welcome contact (if it is, as we assume, on an information gathering expedition), with a civilization capable of detecting its subtle but not secret existence.

So come on, before you start getting all "space-operatic, edge-of-intergalactic-war" paranoid, try to imagine what would motivate a society to create SPRs in the first place, and what they'd hope to learn, and what you think they'd be worried about and might fear, if anything.

How about building self-replicating telescopes instead of probes?

try to imagine what would motivate a society to create SPRs in the first place

Inherent malice? ;-)

One word: bandwidth. Telescope observations produce torrents of data today; if you've got exponentially breeding telescopes flooding the galaxy, then you're going to have real difficulty handling the information flow.

Worse case: your telescope includes a laser transmitter with, say, 1kW output to send a data signal home (that's being incredibly optimistic about receiver sensitivity for long-haul interstellar comms). When you're up to a billion telescopes, that's a terawatt hitting you ...

The idea was not to flood the galaxy with telescopes (it's not different from using probes) but to use lots of telescopes near home as one unbelievably huge astronomical interferometer.

And why would you make them all fire their lazorz back into one point? Use multiple receiving stations. If you can breed telescopes, you can breed receivers too.

Check this out: "The hypertelescope: a zoom with a view"

http://eugen.leitl.org/tt/msg34085.html

Now scale it all the way to Neptune, and you may not need probes at all.

Lignin isn't inefficient. Its the reason the green plants won the race for capturing surface sunlight. Bacteria can't eat them. Dry land was colonised by symbioses between plants and fungi. Later - much later - some of the symbiotic fungi evolved through parasitism to become rotters. The gap between the two accounts for much of the coal. That's why there is a Carboniferous!

Though as others said there was some coal before and since...

But anyway, the whole "peak oil" thing doesn't really matter on the big scale. Life makes the hydrocarbons. There is plenty of time to do it all again and again and again.

Time scale, my dear chap, time scale. If you want to use geologically young coal (<65 myr), it's mostly lignite, which has a lot of water and not much energy density. To use it efficiently, it's better to burn it near where you get it out of the ground, and some deposits would take more energy to extract than you get back out of them.

The higher grades of coal (bituminous coal or anthracite coal) are mostly Carboniferous or Permian in age, although there's nice deposits of K-T anthracite in the Canadian Rockies and Andes (mountains can add extra pressure, apparently). I don't know enough petroleum geology to know if this is a universal pattern, but it does appear that longer times underground improve fuel quality.

So, you're right: all that carbon will eventually go back in the ground. Eventually is probably on a scale of 10,000,000 years, followed by >50,000,000 years of geological forces to turn all that buried carbon into something vaguely usable.

The small problem is that most mammal species have a fossil record of ~5,000,000 years before they disappear, so we're unlikely to benefit from the new glut of fossil fuels.

Still, "The New Carboniferous" is a great title for something in science fiction, IMHO. Society will look very strange in the era when all that carbon starts going back into the ground. For example, the world climate will (likely) look like the early Eocene (I'm really skeptical about us getting our act together to stop climate change).

Personally, I think the Eocene boreal temperate forests of Greenland and Antarctica always sounded really cool. Unfortunately, the billions of people displaced by ~50-100 m sea level rise as the poles melt will cause a few (ahem) minor inconveniences here and there.

Anatoly:

There are no SRPs because it is safer, easier and more efficient to use one solar system as a central factory, build all probes there and send them directly to all stars in the galaxy.

This is the "range" theory, the counterpoint to percolation, predicated on the relative efficiency of probe self-replication and the home system's resources. For energy and matter, they're undoubtedly more convenient at home than abroad. For reliability and range... Difficulty of success at increasing distances eventually demands intermediate "probeyard" locations.

Charlie:

There's a problem with that -- you're assuming that you can build probes with a mean time between failures greater than the the time it takes to directly reach all stars in the galaxy.

If your probes are reliable enough to reach stars c. 150,000 light years away (which at 10% of c would take 1.5 million years) then yes, one solar system will do (as long as you don't mind launching a few hundred billion probes). Otherwise ...

You may do better with a hybrid model: just one star system in a thousand hosts a factory, which probes 1000 nearby systems and curates the report of the results, then identifies an ideal candidate for another factory near the edge of its reliable travel distance and sends a new factory there. In which case only 1/1000 visited star systems gets to host a factory (or 1/100 or 1/1,000,000; it's a trade-off between your engineering reliability and the distance you're surveying).

This also points to a time-domain problem with most of the theories. Ok, so you either spend a few million years sending SRP probes out (with few or many factories, doesn't matter), or if you can build 1.5 million year lifespan probes send them all out from location X. Now you've sent a wave of exploration throughout the galaxy. So, if it was SRP probes you have little factories everywhere. If not, then they can build monitoring posts that don't self-propogate, and keep an eye on things (evolution happens, right? want to take another look every few thousand or hundred thousand years at life-bearing planets...). Or if not, they dissapear as they wear out eventually.

The difference between a SRP system and one that maintains itself is only a matter of degree. It may not have the brains to propogate interstellar distances, but it's communicating long distances (large lasers, presumably) and building lightsail or antimatter or fusion probelets to fly around the solar system and report back on developments (which are similar to interstellar probelets, only smaller and shorter lifetime). Unless you pre-hardwire probelet designs and design in no adaptive behavior it's going to evolve and adapt. And even if it's not that bright, it has millions of years to develop SRP behavior and go out exploring.

You can do all adaptive behavior "at home" but then the success level of the local inspection probes depends on latency and bandwidth home. Challenging local situations equals hard to actually set up shop.

I forsee a very difficult middle ground between probe systems that act functionally like a discovery wave and then burn out, and ones that "last forever" in a SRP system of some sort.

The discovery wave then gives you a complex problem of how long it's been since your last exploration went through a system and reported back, and what may have happened since...

Two things:

1) We don't actually need a fully-general definition of "intelligence" here. Nor do we need a way of measuring dinosaur intelligence directly. All we need is reasonably good agreement on which present-day animals are more intelligent than which other animals; we can then look for things that correlate well with greater intelligence (like EQ, possibly), and measure those for the dinosaurs. If the dinosaurs score worse on all these proxy measures than their modern-day equivalents, it's strong evidence (though not a cast-iron proof) that modern critters are more intelligent.

2) Regarding EQ specifically: according to that Wikipedia article, a possible reason for the increase in EQ is greater innervation of bodily systems. Modern mammals, it seems, are like modern cars: they're full of processing hardware, but it's used for monitoring previously dumb systems rather than for general problem-solving ability.

Interstellar travel is energy-intensive so you want to make the payload as small as possible or leave the engine back home and use beamed power (e.g. a microwave or laser sail). So what we're looking for is a small sail-driven capsule containing a seed...

That's certainly a viable alternative. I guess it depends on how stealthy you'd like to be.

The problem is that lasers and microwaves are fairly obvious to the target civilization. (I certainly hope that if someone was beaming a laser from Tau Ceti to Earth so their probe could use a solar sail we'd notice it.)

The really cool plan would be to aim the laser-sailed probe so it would stop a light-month or so off-center from the target star. That way the laser would hopefully be undetectable to the target system. Then the probe would use Oort cloud material to build the means to get closer to the target. This takes advantage of both approaches; an economical method of probe propulsion and a some very stealthy ways to probe the target solar system.

I'd say we have a winner at this point. (And we know why we haven't see any SRPs.)

That's a very intelligent idea.

The problem goes as follows: One relativistic missile can ruin your whole day. It doesn't matter if 999 out of 1000 alien races are friendly. The single, unfriendly alien can still take out your planet.

I'd assume that an intelligent, compassionate alien race can understand why we'd want to be stealthy. If they're capable of building an SRP, they've sweated the same issues.

The possible relativistic counter-missiles are one of several reasons to have expendable robot factories thousands of light-years from your home system actually doing the exploring. This is not a new idea.

The discovery wave then gives you a complex problem of how long it's been since your last exploration went through a system and reported back, and what may have happened since...

You can just keep on sending exploration waves. It's not like there's lack of matter in the Solar System. Then you get constant feedback while not risking SRP proliferation.

As I showed earlier, mass of Mars is enough to probe the entire galaxy with a huge margin. So, if you can disassemble a gas giant, you don't even have to visit your neighbour star.

Expendable robot factories thousands of light-years away from home won't save you from relativistic missiles hitting your home.

The solution to relativistic missiles is simply living in space.

@charlie: A billion transmitters wouldn't result in a terawatt hitting you, it would result in a tiny change in output from a billion stars.

@anatoly: For serious interstellar imaging performance you want one of these babies!

http://www.tsgc.utexas.edu/archive/design/foci/

Multi AU gravity lens should be able to resolve just about anything. Interesting signal processing problems too.

The solution to relativistic missiles is simply living in space.

The prerequisite for living in space is to stop being human. (Humans are an organism that coevolved as part of a complex planetary biosphere and which are intimately dependant on both the biosphere and the planet -- including dependencies on the planet for such things as gravity, radiation shielding, and scale pressure.)

Approximately speaking, received bandwidth is a function of transmit power. If we want to receive useful signal from 1E9 (or did you mean 1E12 Charlie?) instuments, even after amalgamation and rebroadcasting relays, we still finish up with $home_system getting a TW or so of extra power.

We're talking about laser beams, which by definition are tightly focussed, rather unlike stars which transmit power to an entire sphere. Accordingly power drop-off with distance is proportional to the square of range rather than the cube, so power transmittance over astronomical distances is hugely higher.

Also, we still have to extract meaningful signal from the laser, so we need a useful receive power.

Don't buy it I'm afraid. As you said yourself the inverse square law doesn't magically go away just because you are working with short wavelengths. Sure it is easier to focus lasers than radio, but for any sane mirror/lens size the beam is still going to be huge by the time it gets back to our solar system.

I really can't take the problem of incoming energy from interstellar comms lasers seriously. It's far more likely that you would need dedicated telescopes just to detect it.

So why would any civilisation opt for interstellar probes rather than Eugen's suggestion of hyper telescopes? The latter can return almost all the information a probe can, and can do it instantly for a lot of star systems at a cost significantly less than a single probe.

A telescope, even one taking up a whole solar system, only gets a view from one angle. The probes can get multiple angles, and even see what's behind that nebula by the simple expedient of going and looking.

Ah, you've just invoked the inverse-cube law? Did you mean to do that.

As you point out, the output from a star radiates to the entire surface of a sphere whereas that of a laser can be focused down to a fractional arc-second. But it's still subject to the inverse-square law. If it's a fraction of the surface area of the sphere at 1 AU, it should be the same fraction of the surface area at 1,000 AU, or 1,000,000 AU. Although lasers are (IIRC) a bit weird when it comes to focus issues, I don't think they have anything magical to cause the individual photons to recollimate any time they try to spread. The real advantage of using a laser is that the absolute power for visibility at a distance is that same fraction as against an entire star - you're not wating power by shining in all directions at the same time.

(I assume a point source star for simplicity, which is not the case at 1 AU, but is close enough at 1000 AU and beyond.)

I'm not disputing any of that, except for the bit where you seem to be arguing that there isn't a need for sufficient RX power on the signal you're interested in to make it detectable over the RX noise of $star that it's right next to if not actually in front of.

The very narrow bandwidth of a laser makes it far easier to see against the glare of a star, esp if the wavelength is chosen to be unusual.

I think we're looking at different aspects of the same problem; my points were that for an observer at origin O, noise from $star falls off as Range^3, but signal from $laser only falls as R^2 (presumes R and collimation are such that the entire beam hits system O but we can't sensibly build a receiver that big).

As per #220, the laser has to deliver sufficient power to be discernable from the star rather than just a random piece of black.

Why do you say as Range^3? As far as I'm aware, radiation declines according to the inverse-square law. I'm not sure what your 'noise' is if it's not the radiation.

It could be that intelligent life doesn't bother with probes as they are too much hassle, for all the possible reasons discussed here and more. Rather they look for "intelligent life" within the era where they are biological with telescopes and then go live around a nice singularity as post biologicals where we can't see them any more. Eventually we might find remnants of a civilization, but maybe only as a fossil record. If we get that far we might find that the singularity we have chosen to go live round is already really rather crowded. They might also be hiding from us. I mean have you seen the state of the place? Who wants to talk to or visit the noisy litter louts, let alone invite them round?

Nevertheless, a telescope would be far preferable in terms of cost and immediacy of information. It could survey millions of star systems at the level a probe could manage for (probably) a millionth of the cost. Restricted views would not be a problem unless one wanted to explore over tens of thousands of light years. In which case waiting tens of thousands of years for a signal from a probe might be a show stopper.

The inverse square law is exactly what I'm talking about, but the radiation received by a point at distance R from a star is proportional to the surface area of a sphere of size R, and A =4/3PI()R^3.

Noise is radiation that you're not interested in examining, and signal is radiation that you are interested in.

The trick with collimation and the inverse square law is that collimation creates a "virtual" center point for the sphere far away behind. Intensity at a distance d from the laser output, compared to the intensity at the output, hence depends on the virtual center distance R: I(d) = I(0)*R^2/(R+d)^2

When d is big compared to R, it behaves as 1/d^2. When d is small compared to R, it's way flatter (1-d^2/R^2, top of a parabola).

OG.

A =4/3PI()R^3.

No. A = 4 * PI() * R^2

Your equation is the volume equation. You can sanity check by doing basic dimensional analysis: both 4/3 * PI() and 4 * PI() are dimensionless values, so the difference is whether the result of the equation is R^2 or R^3. If the distance R is measured in metres, R^2 is measured in square metres and R^3 in cubic metres. It makes no sense whatsoever to denote area in cubic metres.

Headdesk

Sorry.

(It's a bit late to disentangle this sub-thread and remove it outright.)

I find the discussion about optical telescopes vs probes and detecting fusion drive probes somewhat blinkered by our current state of technology.

Even if we granted telescopes extraordinary capabilities vs a local probe (I don't), they are most definitely subject to c limitations. It may well be that probes are not and can communicate at FTL speeds. We just don't know.

Similarly with the probe drive. Talk of stealthing a fusion drive exhaust may be like pre-C20th people putting sackcloth on horses' hooves. We have no knowledge of what the main and maneuvering drives of an advanced probe might be.

The prerequisite for living in space is to stop being human. (Humans are an organism that coevolved as part of a complex planetary biosphere and which are intimately dependant on both the biosphere and the planet -- including dependencies on the planet for such things as gravity, radiation shielding, and scale pressure.)

Charlie, if we can create a completely artificial hyper-life (SRP is a living organism that has the entire universe as its ecological niche), then building artificial biosphere would be a piece of cake. And gravity+shielding is stuff we already have solutions for.

That can help, certainly. But at a distance of light years, even 1 AU of virtual centre distance will be lost, and I strongly suspect that 1 AU is some orders of magnitude greater than the best achievable with any foreseeable technology.

a telescope would be far preferable in terms of [...] immediacy of information

This is true but so what. For surveying basic planetary data that doesn't change, much like Kepler, obviously that is true, and useful. But if the survey is also about detecting rapidly evolving technological civilizations, then it is useless. You need "probes on the ground" as it were, to gather intelligence. Of course c limitations gain you little in terms of usefulness, unless you can circumvent this.

But a local probe can do so much more than a telescope, potentially infiltrating and acquiring a lot of information that could never be achieved using remote telescopes. Even simple things, like determining how the local biology works would to relatively opaque to telescopes.

Well, it's pretty useless talking about the possibility of aliens observing us in minute detail via wormholes opened above our homes (or in them).

But a local probe can do so much more than a telescope, potentially infiltrating and acquiring a lot of information that could never be achieved using remote telescopes. Even simple things, like determining how the local biology works would to relatively opaque to telescopes.

Telescope can tell you what systems are worth probing. Then you don't need to send billions of probes, but perhaps only a few, where you found life.

My own guess, and this may speak more about me than about our hypothetical more advanced ETs, is that a civilization curious enough to build mega-telescopes would be at least curious enough to send probes. Not because they have to, but because they want to.

They may have no curiosity at all left for the real world, for example if they went completely virtual. Hypertelescope is a much safer defense system, you can scan the galaxy for threats without risking exposure with your probes.

Even if they could use wormhole telescopes, they couldn't do local sampling or experiments without being able to manipulate local resources. The probe's capabilities are always going to be larger than a telescope's.

Telescope can tell you what systems are worth probing. Then you don't need to send billions of probes, but perhaps only a few, where you found life.

Perhaps. We have telescopes. Can you definitely tell me the status of life on/in the other planets of our solar system, apart from earth?

"Can you definitely tell me the status of life on/in the other planets of our solar system, apart from earth?"

Yes - nothing interesting

Perhaps. We have telescopes. Can you definitely tell me the status of life on/in the other planets of our solar system, apart from earth?

We don't have telescopes that SRP-capable civilization could have.

I can believe there are plenty SRPs undetected in our Kuiper belt. But there is one gaping omission in this article. These SRPs don’t just hang around; they were sent out to gather information and report home. So the problem isn’t just that we’re not seeing any SRPs, were not seeing any message traffic. While this message traffic is likely highly directional, I would expect us to be in the ‘cone’ of such messages (from any solar system). And we aren’t.

But adding messages to the equation poses changes the SRP estimate.

To start with, Fr is assumed to be high. Really? These are advanced civilisations. They have survived nukes, resource depletion, climate change (yeah, assuming..). I would expect them to be risk averse and maybe not so happy putting up ‘here we are’ arrows in the entire galaxy using directional messaging. Fr is probably not so high.

Then the formula assumes any SRP project sends out an SRP to every solar system. Really? Can they send messages all the way home from the other end of the galaxy? Okay, maybe there is a relaying system but that needs not to be assumed. SRPs may well be limited to a sphere or disk of maximum distance from the home world. So the formula would need a factor for a percentage of the number of spheres of SRPs we are in.

Then the formula assumes that these SRPs stay up forever until the end of time or a nova on its target. There are several reasons why they might have an intentionally limited lifespan. Their source civilisations simply might have ideas about not polluting the universe. Or might not want burden or endanger their far descendants with that ‘here we are’ pointer of directional messaging. Or a kill switch might be built in to hide. ‘If you see a nuclear flash, dive into the star’. So lifespan of SRPs needs to be a factor.

So, many reasons why Nr could be much lower.

http://www.space.com/9303-claim-alien-signal-planet-gliese-581g-called-very-suspicious.html

"Looking at one of these objects, we found this signal. We found this very sharp signal, sort of a laser lookalike thing which is the sort of thing we're looking for ? a very sharp spike. And that is what we found."

That's what I said 200 comments ago. I guess I shouldn't have left out the word "hyper," as it is clearly important for communicating this concept.

Relevant to this discussion, an article about seeing probes in the Solar System. We've been littering other parts of the system with high-tech trinkets for a few decades now, and it's not so hard to find them when we know where to look. As pointed out in the article, "But it might be difficult to distinguish between a space probe and a rock."

Thanks for commenting on my paper.

QUOTE [ [ [ [ [ But there is one gaping omission in this article. These SRPs don’t just hang around; they were sent out to gather information and report home. So the problem isn’t just that we’re not seeing any SRPs, were not seeing any message traffic. QUOTE ] ] ] ] ]

Sure, but that just exacerbates the SRP contribution to the Fermi Paradox by suggesting that they should be easier to detect than might otherwise be assumed. I don't see how this is a flaw in the argument presented by the paper since the paper is a thinly veiled defense of the Fermi Paradox in the first place. What you just did was strengthen the paper...in my opinion.

QUOTE [ [ [ [ [ Then the formula assumes any SRP project sends out an SRP to every solar system. Really? Can they send messages all the way home from the other end of the galaxy? Okay, maybe there is a relaying system but that needs not to be assumed. QUOTE ] ] ] ] ]

I have heard this counterargument before and to be honest, I don't understand it at all. Isn't that like saying Columbus never bothered to venture across the ocean because he couldn't possibly report his findings back to Spain? Explorers often travel far beyond the bounds of communication. In fact, that is kind of the definition of exploration if you ask me. Perhaps I am missing your point. I'm sorry if I have misunderstood you.

QUOTE [ [ [ [ [ Then the formula assumes that these SRPs stay up forever until the end of time or a nova on its target. There are several reasons why they might have an intentionally limited lifespan. Their source civilisations simply might have ideas about not polluting the universe. Or might not want burden or endanger their far descendants with that ‘here we are’ pointer of directional messaging. Or a kill switch might be built in to hide. ‘If you see a nuclear flash, dive into the star’. So lifespan of SRPs needs to be a factor.

So, many reasons why Nr could be much lower. QUOTE ] ] ] ] ]

All sorts of speculative scenarios can either increase or decrease Nr. You suggest ways to decrease it, but with just as much whimsy we can increase it. For example, perhaps a typical SRP mission sends ten probes to each solar system to mitigate the effects of probe-failures. I don't see much point in imposing layers upon layers of speculative details upon these scenarios. They don't really inform of reality very effectively since we have little way of guiding such speculation.

I prefer to keep the assumptions to a bare minimum, which admittedly means that the results of such thought experiments are extremely general. It's a typical accuracy/precision tradeoff. You are attempting greater precision in your comments above, but possibly at the cost of less accuracy. I am shooting for extreme generality, and therefore admittedly low precision, but hopefully with greater overall accuracy. My method is imprecise in that it doesn't tell us much about the psychology or motives of utterly alien species, unlike your method, which posits that a significant proportion of aliens dislike pollution (as one example). While my method is not as precise on these matters as yours, it minimizes our imposition upon alien behavior while still giving us something on which to analyze and consider our current observations (the utter vacancy of the cosmos).

Cheers!

Almost 250 posts and not one person, unless I missed it, has brought up the possibility, slight as it may be, that perhaps one or more "UFO" incidents might be a legitimate sighting of an SRP? Come on folks, admit it might even be possible.

I'm much less interested in pie tins poorly superimposed into photographs, and much more curious about a growing body of evidence brought forward by quite legitimate-sounding former military and intelligence personnel who recount stories of objects being sighted over, say, nuclear missile silos, and then the missiles go offline for a number of minutes or hours.

Sure, it all might be misinformation, mistakes, confusion, etc. But I am open to even the possibility that maybe one or more of these situations was legit. All it takes is for ONE to have been legit, and Fermi's Paradox goes straight to heck.

Up until 1940 or so, it would have been possible for a sufficiently advanced probe to deposit cameras, sampling devices, etc. After 1940 or so it would have been too much of a risk, because once we have cameras connected to integrated circuits, radar, advanced chemical analysis, etc., it becomes way too easy for us to discover that we are being probed, and anything left on the ground has to go away.

It wouldn't surprise me a bit if some of the UFO activity in the 1940s and maybe even into the 1950s was the final pick up of sampling/camera devices.

Its already been shown that energy can be made and stored in "black ponds." A large pond with a black bottom, filled with brine will make and store solar power. The thermal mass of the pond will store the heat and a low temp. steam power plant can make electricity as needed. Add power lines from alternate energy going to heaters and power can be stored and used very well.
Radio would signal Intelligence. We have not heard any. Maybe they are some kind "FM" and we are AM. And maybe they have been listening to Rock and Roll for a few years. What will they think when Rush gets there and starts beaming in.

Hmmm... all those stories about aliens probing people. Maybe those were self-replicating probes... 8-)

Almost 250 posts and not one person, unless I missed it, has brought up the possibility, slight as it may be, that perhaps one or more "UFO" incidents might be a legitimate sighting of an SRP?

This would suggest considerably more on-site intelligence than what we've been modeling here. Two points arise when looking over "UFO" reports:

(1) Events are transient, with little or no lasting evidence. We might get blurry photos or oddly scorched fields, but no robotic rovers exploring London.

(2) Explanations presented generally fit the world-view of the humans involved. So we get flying saucers from Venus, airships built by mad scientists, appearances of the Virgin Mary, encounters with the Fair Folk, and so on. Some of these explanations seem dubious to others.

Assuming for argument an outside agent causing all of this - rather than something poorly documented in the human nervous system - we are left with the conclusion that whoever or whatever is doing this does not want open and unambiguous contact. They may want more subtle things such as psychological experimentation, plausibly deniable social engineering, or just yanking our chain "for teh lulz." As evidence of extraterrestrial robots - or time travelers, or angels, or what have you - it seems rather poor.

If and only if the radar would detect the probe, the camera was not made of materials that would pass for indigenous etc etc. It's at least equally possible that these events were the surveillance network being installed in a window of opportunity before we got the chance to build the globar radar surveillance networks etc.

Is there any advantage to sending large telescopes out as “probes”?

Any probe would be a telescope as well, obviously.

What we are talking about is probes vs. enormous hypertelescope composed of millions of telescopes working together

Remidns me of Schlock mercenary's Very Dangerous Array

Proactive Target Acquisition.

Awesome.

I was thinking of sending a number of very large arrays some significant distance from the home system to get a different perspective on the things being looked at.

Well, once you have self-replicating probes, you can damn well build anything you want anywhere you want. ;-)

Or maybe ET has found a way to efficiently detect neutrinos and uses neutrino beams. No problems with being blocked.

That could also be the case.

There are "Street People" in every city on the planet, talking to "themselves", wearing bulking layers to hide their shape. They are obviously alien anthropologists monitoring our every action, and constantly reporting back home, all in plain sight.

UFOs would not land on the White House lawn to meet with lying sociopaths, or get shot by the security service. They would come among us to see how we treat the least of us.

Plus, don't forget all those muttering IT people that we routinely ignore. The story "Antibodies" was clearly nonfiction. HA!

Yes, they would come among us. But probably like an advanced form of this: http://www.youtube.com/watch?v=_i-_1QdY2Zc&feature=related

Synsects? Certainly, and great espionage technology until we actually find the first one. I think the CIA has attempted to do this since the 1970s.

I've been having a lot of fun writing about time travelers using synsects as microdrones. One of my favorite ideas is "Horsefly Productions," the company which recorded the entire works of Shakespeare at the Old Globe Theater, using a composite of the flies' eye and moth's eye views. The trick for them was blending in to Elizabethan London long enough to get the recordings.

Of course, time travelers also recorded Jesus, Buddha, and Mohammed with the same technology, which gave the travelers a rather different take on the world's major religions.

A sufficiently advanced technology would make then almost identical with the real thing. Maybe only the bio-engineering of their brains would give them away, and how often do people check insects to that level?

I think it would only take one find to give the show away. Besides, someone would eventually notice if the bugs around them were radioactive (in the sense of giving off some sort of radiation that they use for communication).

Yes, they could be using neutrinos or dark matter or whatever untraceable exotic methods for communications. However, if you want me to believe that there are synsects out there that are microscopically identical with real insects, and that use some untraceable communication channel, I'll just claim the roaches are psychic and go put on my tinfoil hat.

"I think it would only take one find to give the show away."

So, you have to find one out of trillions. A rather improbable task given that distinguishing between the natural and synthetic would likely take an electron microscope.

Comms is another matter. If they were using, for example, terahertz radiation our tech is not yet up to detecting it except in a lab setting.

There's a fair amount of research going on about the bugs that live near us (cockroaches, bedbugs, bees, flies, etc), usually figuring out how to kill them better. The roach that refused to die because it was an alien robot...that might stand out a bit.

Similarly, a CIA dude sweeping a place for bugs and finding an otherwise unremarkable silverfish that was transmitting alien radio code, that might also stand out a bit.

Since Earth's atmosphere is largely opaque to terahertz radiation (unless Wikipedia is lying again), it's not a great medium for sending a message into space. I'm not worried about it. Nor am I worried about insects being little ELF broadcasters either.

I’m not introducing signalling as a way to increase the visibility of SRPs, I’m just assuming signalling is detectable and directional and could therefore reveal the source of these SRP’s.

Now your formula is relevant to us n00bs doomed to sit on their rock and look up for life. For those starting an SRP mission a different formula will be relevant. This one:

F * U * C * K * E = D

Where F is the number of alien societies you will encounter, U the fraction that is aggressive, C the fraction that possesses planet busters that travel at near light speed (or technology to similar effect), K the chance that they will launch a mission, E the chance that they will find you and D the number of times your planet will be destroyed. Setting any of these variables to zero will be a dangerous assumption. Add some math to F describing the expanding wave front of your SRPs and the galactic distribution of alien life and you can calculate you chances of making into next year.

It’s not about the number of assumptions you make, it’s about making your assumptions explicit. You have implicitly assumed that most capable of building SRPs perceive the galaxy as benign. I’m saying: not so fast! So I think your Fr is too high, or in order to mitigate risk, reduce F in my formula and send out a mission limited in time or space and add something to your formula to express that not every SRP mission goes everywhere, forever.

I'm not sure you can't get round this fairly easily unless you assume that $hostile will launch an invasion rather than an automated planet-killer.

Build in some relay stations, and when one of them goes quiet, you know the approximate direction of $hostile, while if they haven't obtained the direction of the transmission that the relay was sending in, they only know that they've killed one planet. Even if they do have the TX beam from the relay, you should have at least as much reaction time as they do, unless they've used packs of planet killers, and have sent at least one more killer than you have relays in the chain.

True, but would you bet your planet on that? You stand the best chance of not getting found if there is no on looking for you. All I want to say is that risk perception might explain the Fermi paradox.

There's another factor I think you may be missing: The problem only occurs if $hostile will try and kill anyone who finds them, rather than anyone that they, say, detect radio band signals from. If they will try and kill anyone they detect radio band from, you're probably too late to stop them by the time you start having this conversation, so you urgently need to find someone friendly with planet killer hunter-killer tech.

Despite the unnecessary flippancy displayed in your post I will briefly entertain your primary concern that my original formula didn't specifically include a parameter to account for limited probe lifetime once reaching its assigned star.

Personally, I think it's preposterous to render a super-advanced super-intelligent being with absolutely unlimited control over matter and energy down to at least the atomic level anything short of utterly immortal (aside from transient events such GRBs or supernovas, and those can, to some degree, be forecasted and shielded against...although this is an interesting idea that is worth pursuing, more interesting than intentional apoptosis if you ask me; I'll try to run some numbers on it.).

Your point is that these beings have no choice since it was rudely imposed upon them by their cruel creators. This gets to the heart of our disagreement. In my mind, SRPs are not just third party "tools" of the "real" society back home; they are the rightful and complete members of the original society in and of themselves, leaving the homeworld to explore and colonize the galaxy. If you attained the advanced technological capability to fly through space and explore the galaxy for thousands to millions of years, would you include in your own design your own inevitable and eventual death? This is why I find the idea so ridiculous. I don't see SRPs as dumb tools with no self-awareness, or desire, or will, what have you. I see them as full conscious beings...why would they design their own death into the system? We don't die because we were designed to die by a purposeful agent (a lot of people would disagree), we die because evolution is blind to such concerns in the first place. Intelligently invented artificial beings are not likely to include such restrictions, especially if they gain access to and control over their own systems (as humans are desperately endeavoring to do with modern medicine).

People insist on viewing SRPs as mindless external third party tools. I tried to undo that bias in the paper, but it's a really tough concept for people to get around.

If you want SRPs to die with no say in the matter (like dumb tools) in your personal vision of galactic exploration, go for it. That's what we're all doing here: honing our own hypothetical scenarios of an incompletely observed system to fill in the gaps as best as we can. In your personal view of super-advanced societies -- be they aliens far ahead of us, or alternatively, we ourselves in our own distant future -- computerized intelligence is second rate, something to be controlled and manipulated by the "true" members of society. In my mind it isn't and would be unlikely to design its own death into its own future. That's the distinction. You are absolutely correct that including such an assumption and its associated parameter(s) into equations intended to estimate the SRP population will doubtlessly reduce the number SRPs estimated to be present at any given time. I can't argue with the mathematical effect of adding such a parameter to the equation, but I can argue with the philosophy underlying its justification.

Yay!

I will note one more thing. If an alien race is really, really F U C K E D, they will not just destroy the probing race, but ALL races the probing race has discovered.

This is why Keith's paper needs some serious revision. I'm seeing three errors that stand out above the rest. Before discussing them, I should note that Keith's objections against second guessing an alien mindset are mostly valid in my opinion, and they are why I'm not throwing the kitchen sink at him in terms of criticism. (The 270 comments above make not only a kitchen sink, but a dish drainer full of dishes, a towel rack, and some leftover chicken soup available for critical hurling.)

THE FIRST is your F U C K E D equation above. This translates into the need for concealment. Leaving aside the possibility that another race might destroy you, having your probe discovered is not necessarily a good way to begin diplomatic relations. So Keith's paper must discuss the issues of stealth, concealment, and the very good reasons for same. The assumption that we can easily detect a probe is just plain silly for a variety of reasons.

I am aware of my argument's implicit assumption that an alien ecology would produce concealment. Given that our aliens have senses of some kind, and that random mutations in genetic design (whatever aliens use for genes) would eventually lead to organisms which are harder to sense, the burden of proving that aliens from a complex ecology don't know about concealment lies upon someone who wishes to assert this as fact.

SECOND is the issue of mission planning. I'd assume that an alien race who could build SRPs is capable of developing a plan which would prevent duplication of effort, arrest the infinite growth in probe numbers, (with the increased chance of discovery) and prevent probe mutation. (Less probes = lowered chance of probe mutation both through less total probes and fewer generations of probes.)

THIRD is the issue of time. We've been actively scanning the skies for a very short time historically speaking, and we've had the necessary coverage to make it probable we would discover an unconcealed probe for only fifty years or so. When I'm told we haven't seen a probe in fifty years (ONLY fifty years!) my reaction lies someplace between "So what?" and "Duh!" The need for concealment and the obvious advantages of mission planning make it highly unlikely that we'll see a probe anytime soon.

If we don't see a probe in the next thousand years I might apply the data to the Drake Equation but for now I regard the entire issue as insignificant.

You are essentially correct. Obviously a Self Replicating Probe is capable of repairing itself, thus it is more-or-less immortal. However, if I was involved in mission planning, I'd arrange for each probe to stay in a single solar system as long as possible, and only leave in the event of probable discovery or grave danger. In other words, a probe might stay in a single solar system for billions of years.

Essentially, as a probe, you reproduce once and then become sedentary.

I will note one more thing. If an alien race is really, really F U C K E D, they will not just destroy the probing race, but ALL races the probing race has discovered.

Also note: if you can build self-replicating probes with signalling devices capable of being heard across interstellar distances, then you are probably capable of manufacturing a Dyson sphere (of the free-flying solar collectors with computing gizmos attached). And if you can do that, you can build a Dyson-Nicoll laser -- link goes to a gamer resource: I don't think anyone's done any actual quantitative work on this Strangelovian doomsday weapon.

Or even of sending Von Neumann probes to a bunch of uninhabited neighbouring stars, building Dyson spheres around them, and coordinating them as a massed battery of Dyson-Nicoll lasers. Which would also have the added unhappy capability of acting as a very long baseline interferometer with each element having a light-gathering aperture on the order of 1AU in radius, spaced a few parsecs apart.

So: anyone capable of building SRPs is probably capable of building an imaging system that can read a newspaper at kiloparsec ranges, coupled to a death ray that can vapourize a terrestrial-sized planet anywhere in the Local Group of galaxies in a matter of minutes (when the supernova-intensity wavefront arrives).

We're all doooooooomed ...

I think you place too much weight on consciousness (in the human sense) as a desirable attribute for a post-human interstellar civilization (which is what a network of SRPs amounts to). See also Karl Schroeder's "Permanence" for more on this matter, or "Blindsight" by Peter Watts (yes, the hard SF authors are ahead of you).

We're all doooooooomed ...

I wouldn't go nearly that far, but if we ever meet a race that doesn't bother to conceal their probes, I'll have some interesting questions about their sanity.

Or you may conclude that the probe is a behavioural probe, designed to elicit a response and determine if we're safe to allow to live. And if we fail the test, a star-pumped phased array laser blast will be along in a few years ...

Charlie, your optimism knows no bounds.

For myself, I think a certain level of concern with security is appropriate, but at this point we're getting too close to "destroy them in case they might want to destroy us" territory. I'll settle for carefully concealing our probes and dealing carefully with newly discovered aliens.

Of course, the Great Filter could look something like this:

Each race, before building their SRP fleet, puts some kind of advanced retaliation device into service, much like the ones built by Moscow in Iron Sunrise. So far, this makes a certain amount of sense. Each probe, upon being discovered, announces that their creating race has a relativistic retaliation capability before opening diplomatic relations/running away/self destructing. The general hope is that this will induce F U C K E D races to refrain from conducting xenocide. (The SRPs could even be the retaliation devices, though I'd keep the functions separate.)

Unfortunately, there exists just one F U C K E D race that attacks anyway. The people from the destroyed planets launch their retaliation devices against whoever they might be paranoid about, and a chain reaction ensues... Within a few thousand years, there are no intelligent races in this area of space, and any survivors aren't launching new probes or communicating with anyone at all.

Alex, thanks again for the continued discussion...

Sorry for the length...

QUOTE [ [ [ [ [ THE FIRST is your F U C K E D equation above. This translates into the need for concealment. Leaving aside the possibility that another race might destroy you, having your probe discovered is not necessarily a good way to begin diplomatic relations. So Keith's paper must discuss the issues of stealth, concealment, and the very good reasons for same. QUOTE ] ] ] ] ]

I don't see how this relates to my paper. It wasn't specifically speculating on the ease of detecting such probes (their likelihood of being concealed), I was estimating the number that exist, regardless of detectibility. In fact, at the end of the paper I specifically advocate more aggressive probe-search SETI efforts. If by concealment you mean that due to DharmaBum's model no one would attempt such a mission in the first place, then that's not probe concealment, it's avoidance of the mission entirely. Fair enough, we can consider that, but my usual argument would apply (I feel like I sing this refrain a lot): how can such a consideration be universally applied to every species to ever arise in the entire galaxy? It would take but a single gregarious species out of presumably tens of thousands of unfathomably diverse alien species to rapidly populate the galaxy with loud-mouthed SRPs. Where are they? The argument that, once discovered, they are subsequently wiped out is specifically shown to be erroneous in the paper in my response to Chyba and Hand (section 2.3). They physically cannot chase down and eradiate the SRP population as it spreads across the galaxy.

Furthermore (and I admit that we are venturing into borerline fictional musing at this point), I think it's practically a reversal of logic to assume that species B (the visited) must necessarily succeed in destroying species A (the visitor). The more logical conclusion is the exact opposite, that species A is more advanced than B. After all, only species A has demonstrated the technological capability to traverse vast reaches of the galaxy. B, by definition, didn't do so. They waited and struck back at those who came swinging by. By any reasonable logic, A can only be assumed to be the more advanced and stronger species (they're the only one of the two to have proven their interstellar capability while B's interstellar capability is, at best, speculative)...yet we would assume that B, having received A's probe, would be successful in sending a relatiatory strike?...their first interstellar mission ever no less? Are we confident this scenario actually makes sense?!

In fact, I have considered the notion of interstellar invasion and war-mongering in the past (not in this paper I admit, we're venturing into new topics here) and I have always found this notion extremely wanting. An interstellar payload is pretty limited in its capacity (matter or energy, take your pick) relative to the matter and energy resources available to a homeworld and its host solar system. The homeworld has whole planets and asteroids worth of material at its disposal and has the entire fissal and solar energy reserves (and possible antimatter capability) of its solar system at its disposal. Interstellar distances should render any invasion quite impoverished by comparison. After all, we're considering civilizations that are bordering on Kardashev-II here, not Earth. They have their entire solar system's resources to defend themselves with and an attacker must attack with some paltry interstellar delivery. I highly question the notion of interstellar war. The defenders have a huge advantage in any such scenario.

It's like two castles separated by a few miles of terrain, trying to attack one another, but for some reason they can only send an archer or two per attempted conquest. The defenders never lose.

Last thought: this theory of limited interstellar war capability is very similar to the ITB theory in the paper. Interesting. I'll have to ponder that.

QUOTE [ [ [ [ [ The assumption that we can easily detect a probe is just plain silly for a variety of reasons. QUOTE ] ] ] ] ]

Same thing as above. I never specifically addressed the observability of the probes, merely their expected population. I think the part where we are talking around one another is my claim that high SRP populations suggest observability where you imply that they may all be concealed; thus, you are suggesting that I am suggesting :-) easy observability. I think that's your point. My response is the same as it has been throughout this discussion: a concealment argument must apply to all SRPs, it must be a vital motive of all SRP mission designs. This is a strong statement about the behavior and values of aliens. It takes but one species with a completely different approach to detectability and their SRP is landing on the white house lawn to say "howdy". That is the crux of the SRP contribution to the Fermi Paradox. I can grant you 99% correctness in the theory of concealment and it still fails to resolve the Fermi Paradox...unless the estimated population is so low that arguments by numbers fail (if there are only a handful of SRPs in the solar system then perhaps they all -- by chance -- adopt a concealment strategy). That was the purpose of estimating the SRP population in the paper, to illustrate that it may be a pretty high number and therefore problematic for the Fermi Paradox.

All of that said, I also consider (in the paper) the possibility that maybe, just maybe, no openly communicative SRPs arrive. I do give that possibility honest consideration at the end of the paper when I explicitly advocate more aggressive SRP SETI searches.

I feel like we aren't communicating clearly on this. I keep making the same point over and over: that in various counterarguments I see strong assumptions about the behavior and motives across a wildy heterogenous population of alien species and their technology...without the possibility of exception. I don't mean to sound obstinate, but I'm surprised no one understands my point on this. I'm sorry we're arguing so aggressively about this.

QUOTE [ [ [ [ [ SECOND is the issue of mission planning. I'd assume that an alien race who could build SRPs is capable of developing a plan which would prevent duplication of effort, arrest the infinite growth in probe numbers, (with the increased chance of discovery) and prevent probe mutation. (Less probes = lowered chance of probe mutation both through less total probes and fewer generations of probes.) QUOTE ] ] ] ] ]

I completely agree. Most of your points above are made explicitly in the paper (for example, I suggest that SRP missions target only one SRP per star and I suggest a stunningly low mutation rate). Why then do you imply that these ideas are your second challenge to the paper? They're in the paper.

QUOTE [ [ [ [ [ THIRD is the issue of time. We've been actively scanning the skies for a very short time historically speaking, and we've had the necessary coverage to make it probable we would discover an unconcealed probe for only fifty years or so. When I'm told we haven't seen a probe in fifty years (ONLY fifty years!) my reaction lies someplace between "So what?" and "Duh!" The need for concealment and the obvious advantages of mission planning make it highly unlikely that we'll see a probe anytime soon. QUOTE ] ] ] ] ]

Sure. Freitas suggested this in the 80s (and I cited him for it). Freitas suggested, and I mirror his thoughts in the paper, that perhaps we should explore our solar system for such devices.

That said, I do think it's rather odd that not a single gregarious species have ever arisen in the galaxy (if they had, they would have filled it long ago and I have subsequently argued against notions that they could be wiped out after the fact). I find that particular point of absence to be spectacularly bizarre.

I absolutely agree with this comment...except that the replication factor might be slightly higher than one (one offspring). After all, a replication factor of one would never increase the population. If you sent out five initial probes, you would only have five lineages attempt to populate the galaxy.

I think a logical SRP mission would produce enough offspring to populate all stars within some fixed distance (effectively "direct neighbors") and in some subtended cone directed away from the original homeworld, perhaps as wide as a hemisphere. For example, Earth is surrounded by about 52 stars within 15 lightyears (many of which may not be "interesting"). Considering a partial proportion of interest and a partial cone of coverage, this would suggest a few tens of offspring, or even fewer.

Whoops, when I said a replication factor of one wouldn't increase the population I meant it wouldn't do so superlinearly (obviously any replication increases the population), the implication being that doing so linearly is horribly inefficient (that comes back to my original complaint against Bjork and Cotta/Morales in the paper).

Sorry for the poor wording.

Doesn't that approach leave a rather large signature visible from a fair distance for a long time? Ok, not everyone would be near the beam area, but the effects needed to make it might be a bit obvious.

Or maybe you've just worked out what some gamma ray bursts are?

A more stealthy method might be the old comet from Oort cloud one.

I have considered the notion of interstellar invasion and war-mongering in the past (not in this paper I admit, we're venturing into new topics here) and I have always found this notion extremely wanting. An interstellar payload is pretty limited in its capacity (matter or energy, take your pick) relative to the matter and energy resources available to a homeworld and its host solar system.

The twin problems here are "How do you stop a relativistic missile?" and "How do you know the relativistic missile comes from where you think it came from?"

Even with substantial technical differences between the shooter and the target (the target's tech base might be much better) stopping a relativistic missile is damn-near impossible. Also note that per Pellegrino,* a relativistic missile doesn't have to be very large to do considerable damage.

The problem I have with your efforts is not so much the issue of how you deal with concealment, detection, mission planning, etc., but that after doing so, you imagine that it is possible to reach a conclusion. The problem with SRPs as an adjunct to the Drake equation is that we do not have the ability to quantify them at all.

*Pellegrino considered the idea that a relativistic missile can be aimed at a planet's atmosphere, and the impact of a relativistic missile against a planet's atmosphere will cause enough radiation to sterilize both the surface and hundreds of feet below the surface. He suggested that such missiles could very small. Charlie's "laser farm pushing a coke can" would be a sufficiently powerful launch mechanism.

Despite the unnecessary flippancy displayed in your post I will briefly entertain your primary concern that my original formula didn't specifically include a parameter to account for limited probe lifetime once reaching its assigned star.

Personally, I think it's preposterous to render a super-advanced super-intelligent being with absolutely unlimited control over matter and energy down to at least the atomic level anything short of utterly immortal (aside from transient events such GRBs or supernovas, and those can, to some degree, be forecasted and shielded against...although this is an interesting idea that is worth pursuing, more interesting than intentional apoptosis if you ask me; I'll try to run some numbers on it.).

Your point is that these beings have no choice since it was rudely imposed upon them by their cruel creators. This gets to the heart of our disagreement. In my mind, SRPs are not just third party "tools" of the "real" society back home; they are the rightful and complete members of the original society in and of themselves, leaving the homeworld to explore and colonize the galaxy. If you attained the advanced technological capability to fly through space and explore the galaxy for thousands to millions of years, would you include in your own design your own inevitable and eventual death? This is why I find the idea so ridiculous. I don't see SRPs as dumb tools with no self-awareness, or desire, or will, what have you. I see them as full conscious beings...why would they design their own death into the system? We don't die because we were designed to die by a purposeful agent (a lot of people would disagree), we die because evolution is blind to such concerns in the first place. Intelligently invented artificial beings are not likely to include such restrictions, especially if they gain access to and control over their own systems (as humans are desperately endeavoring to do with modern medicine).

People insist on viewing SRPs as mindless external third party tools. I tried to undo that bias in the paper, but it's a really tough concept for people to get around.

If you want SRPs to die with no say in the matter (like dumb tools) in your personal vision of galactic exploration, go for it. That's what we're all doing here: honing our own hypothetical scenarios of an incompletely observed system to fill in the gaps as best as we can. In your personal view of super-advanced societies -- be they aliens far ahead of us, or alternatively, we ourselves in our own distant future -- computerized intelligence is second rate, something to be controlled and manipulated by the "true" members of society. In my mind it isn't and would be unlikely to design its own death into its own future. That's the distinction. You are absolutely correct that including such an assumption and its associated parameter(s) into equations intended to estimate the SRP population will doubtlessly reduce the number SRPs estimated to be present at any given time. I can't argue with the mathematical effect of adding such a parameter to the equation, but I can argue with the philosophy underlying its justification.

I suggest that having human-level sapience as a default is not optimal for the probes' primary mission of looking over the galaxy.

There are only a few times when having a human level intelligence at hand will actually do something that automation cannot. Evaluating 'yes, this is interesting' versus 'no, that's a rock like the last billion rocks' takes a mind (although first filtering can probably be automated), but slow orbits through empty space and waiting millions of years to see if anything newsworthy evolves are best left to computers and dumb automation.

I propose that a human-level intellect is not needed once a probe has arrived, made its initial system survey, and if necessary set up a production facility. (These steps may or may not be practical to automate in full.) So a useful method may be for a small dumb probe to decelerate and take up orbit in a system, then either boot up or download one or more human-level minds. Any on-site thinking that needs to be done in less than the message turnaround time from headquarters can be done; once everything is set up, I'd expect the scout minds to send at least their incremental memories back to headquarters. Not the home system save in the earliest missions, presumably, but to a local factory/colony node. Such a node might well look like a machine mind civilization, hopefully nicer to its inhabitants than some in fiction, although presumably all virtual.

This scheme has the advantages that the probes are fully expendable most of the time, that sapient minds aren't isolated from their fellows more than necessary, and that a robot probe can call home with messages such as 'I've got something odd about this methane covered ice moon. Can you send biologists?'

Obviously, I'm assuming that nothing much comes up that won't tolerate a few hundred years of back-and-forth messaging. In nature, this is generally correct. If the system has residents who are already playing with radio, headquarters will notice as soon as it points an antenna at the target star and the probe can abort its mission (into interstellar space or into the star, as convenient).

Sorry about the double post, please disregard #286, it's a duplicate of #273. I'm not even sure how that happened.

Charlie, are you able to delete it?

Keith: Despite the unnecessary flippancy displayed in your post

Keith, this isn't just a blog; it's a curated online community with a large number of regular visitors who interact with one another in various threads, on an ongoing basis. Flippancy and ironic humour are all part of the rich tapestry; for example, you might want to dip a toe in the discussion immediately after this one (go to the top of this page and click on "Mercury, Retrograde") before you criticise.

If I was designing a probe I would build it in layers.

A autonomous layer built from a very sturdy RAD hardened substrate. Makerbot instructions on how to rebuild higher level structures within itself and boot that with a higher level framework and supply background knowledge (if necessary).

A higher level substrate that clocks at a variable rate, from full on real-time intelligence down to a 1,000 year trip in an hour of subjective time.

Instructions on how to build an advanced hyper-intelligent structure if the situation warrants.

In other words, it doesn't start out necessarily as intelligent as us, but can rebuild into something that is if needed. A fleet of quiet probes finding nothing worth booting into higher levels for spanning the Universe.....

You know, I liked that strategy, until I thought about it some more. AFAIK, the problem with all space travel is that it takes the old military saying about long stretches of boredom punctuated by moments of dread to the nth degree.

The problem with a boot-up intelligence is that it responds to slowly to that whole class of space emergencies that happen very rapidly. Yes, most of these deal with strikes by various small objects at high speed, but the problem is that with a bootable system, the system is trying to gear up at the same time that it's under extreme stress. It's damage control right out of bed, so to speak.

The other end (long stretches of boredom) isn't much better, at least for something that looks like human intelligence.

It seems that the most sophisticated part of such a system has to be whatever is controlling the variable clock rate. However it is done, it has to be able to crank the problem solver from very slow to very fast (or vice versa) very quickly and with little error.

That's a neat technical trick.

I suspect the smart design would be one where replaying the cognitive journal and musing on correlations -- consciousness -- is reserved for the slack time during interstellar cruise, while emergencies and/or arrival in new star systems calls for trained or autonomic reflexes with no conscious supervisor (consciousness being too slow to be useful).

Perhaps there's a black swan technology just out of our current grasp due to our incomplete understanding of physics that allows sub Kardashev I civilizations like ours unlimited but restricted colonization.

Wormholes to another dimension, or the typical jump drives of science fiction enable "high bandwidth" population shifts while absorbing all the effort for real space communication and exploration.

Entirely blue sky, I know, but a little less sinister than the Malthusian relativistic planet killers.

Indeed. The probes don't need self-awareness for looking over the 50,000th empty lifeless system; by the time a replicating interstellar probe can be built, a basic survey should be old hat and fully automated.

Consciousness is only needed occasionally, for questions such as 'should we build a base in this system or elsewhere?' and 'are those critters talking to each other?' - we're already building robots which can do routine things with very little abstract thought.

Yes, I agree that reflexes would be good, in some circumstances. Repair between the stars is a good place for lower level processes, true.

However, I'm not sure about entering a new star system. Aside from the potential issue of sneaking up on a hypothetical new intelligence, there's simply more stuff closer to a star. This means more navigation issues, more potential resources, and more things to investigate. This seems to be a time for the intelligence to be fully active, processing information and setting priorities.

You know, I liked that strategy, until I thought about it some more. AFAIK, the problem with all space travel is that it takes the old military saying about long stretches of boredom punctuated by moments of dread to the nth degree.

The problem with all space travel is that with all foreseeable levels of technology optimizing for one characteristic drastically curtails the expression most other desirable characteristics. Case in point is something we see right now: advanced propulsion technologies for exploring our own solar system aren't being actively developed because minimizing trip times aren't high on the list of priorities when it comes to designing a mission.

This applies in spades to interstellar distances, where slower and smaller beats bigger and faster hands down every time. Especially given that our hypothetical probe has the ability to reproduce. Suppose you have something that weighs on the order of a gram and that it takes 100 days to double it's mass once it reaches the target system.

After 1,000 days, the package masses a kilogram. After 2,000 days, it's up to a metric ton. And after 3,000 days - less than 10 years - over a thousand metric tons. That ought to be big enough to support some serious cogitation :-) And given that trip times are most likely to be measured in centuries at a minimum, and given the amount of energy needed to travel at even 1% of c, it's a far, far better use of your budget to launch a bunch of grammies in the hopes that one of them succeeds as opposed to comparatively few kilogram-sized or larger models.

The "yes but" in sending out gram-sized probes is that it's not clear how much information they can take with them if they're that small.

At the minimum, we want them to know how to phone home and how not to turn into a galaxy-ravaging monster. It might also be good to have some protocols for how to settle a solar system, with things like mapping the gravity manifold, how to identify another intelligence, and so forth.

Can that be packed into a few grams, along with all the other stuff that needs to be there?

I think the big question here is how small a spore you can launch and still have it reconstitute all the probe functions you want at the far end. So as a BOTEC lower-limit calculation, how much information needs to be encoded in our VSP (Very Small Probe)?

That is, how big is our hypothetical probes genome?

Something like what you need to specify a human is something on the order of a gigabyte - maybe more given the cheat that a lot of that information is not actually stored in the germ cell's DNA, but almost certainly not a terabyte's worth (unless I'm badly behind in my ingestion of pop biology articles for the lay public.) So what then? Maybe a petabyte?

If that's the case, then Charlie's memory diamond scheme gives us a mass of roughly 10^15/10^23*10, or roughly about a tenth of a microgram. Throw in the other machinery that does what chloroplasts, mitochondria, ribosomes, golgi bodies, etc do for Earthly biology. You could do that and more with a mass budget of a milligram or so, tops. That's a pretty small probe! Particularly when you consider the fact that after unpacking at the far end the machinery may mass several thousand metric tons.

With probes these tiny, you also don't have to develop terribly exotic space propulsion systems. How high a gee-load can a small mechanical probe safely take? Maybe ten thousand gees? If that's the case, and you want a mission velocity of 1,000 km/sec (0.3% of c), we're talking about a track length of . . . 20 meters. And if the probe masses one milligram, we're talking about energy requirements on the order of a megajoule. Even plain old chemical energy is up to that.

So in this scenario, all you need is a moderately developed linear accelerator - something we could certainly put together by the end of this century, tops. And very likely much earlier than that.

No, small and slow is the way to go, and exotic space propulsion schemes aren't the technology you want to throw money and brains at.

Ah, heteromeles, I think our posts at 19:08 and 19:11 crossed. Hope that's a partial answer to your question.

Is your WAG about the information requirements substantially different from mine? If not, gram-sized probes seem immanently feasible.

10,000 gees is easily survivable by microelectronics -- that's within spitting distance of conventional tube artillery acceleration today, so the existance of precision-guided artillery would tend to substantiate that.

However, 10,000 gees for one second only gets you up to 100 km/sec -- you missed a metre/kilometre conversion in your calculation. You need to sustain it for 100 seconds to reach 3% of lightspeed, which would require a launch track half a million kilometres long (assuming I'm not dropping any decimal places in my head -- bad bet!). So it's probably something you'd do with a photon sail of some kind and a buttload of high powered lasers. Then you run into reflectivity/absorption issues; it's a lot safer to take your time over accelerating the probe and not risk burning a hole in the sail!

Yeah, I did it in my head, but I was going for 1,000 km/sec, or 0.003 c equal 0.3% of c. Hmmm . . . yep, forgot to square, 10^6 goes to 10^12, and 10^12/10^5 is 10^7 meters or ten thousand kilometers. A little bit tougher.

But this is still eminently doable for a linear accelerator in space, where the elements of the machine don't have to be contiguous. I'm thinking along the lines of maybe ten thousand carefully aligned segments floating free in space, with each segment maybe a kilometer long. Note also that we're talking about physical forces which never exceed a few hundred newtons if your ten thousand gees is applied to gram-sized or smaller objects; iow, you don't need particularly massive elements or particularly exotic materials to make the thing.

A lot tougher than I originally made out, but I suspect it's still easier than other advanced propulsion technologies.

Or maybe not; laser pushed light sails work just fine over these comparatively short distances as well, say to no more than a couple of dozen AU's. Assuming top speeds << than c, of course.