Personally, I think that soon, universities are going to look a lot like they did in the 1930s (via post oil). But whatever.

Memorization is like a muscle. The more you train it, the more you can do with it. When I was a TA, I was already running into students at a Tier 1 school who had never taken a foreign language, never learned to memorize anything (a score for music, language, nothing), and who flailed miserably. Their computers didn't help them, because they didn't even know what questions to search for.

The problem is that much of the natural sciences is language (to be precise, it's technical jargon). It's descriptions of worlds that most people never see. If you want to live in those worlds, you need to learn to speak the language. Of course you can let Google translate everything into the few terms you already understand, but why cripple yourself, restricting your mind that way?

As with all the other endangered languages in the world, there are many fields in the natural sciences where there are <100 fluent speakers. A few budget cuts, and a language dies. This can be things like, oh, Drosophila identification (all 800 species, not just D. melanogaster), worm identification (everything from parasites to earthworms), fungus identification (the poison control center has to have someone to call, after all, and the molecular biologists still think Lotus is a water lily), and so on. Worse, there's an inverse relationship between the number of people studying a group and how common it is in the wild, which means that most species are in danger of becoming unidentifiable every single generation.

As for SF...the best way of learning is (as it always has been) small group tutoring, preferably done over tasks that actually apply the knowledge that's being conveyed. That's not going to change any time soon, because a tutor can actually tell when the student doesn't understand, and change the lesson to fit a single student. Lectures have always been a suboptimal solution. The big problem right now is that things like irrational cost jumps and idiotic market demands for degrees are skewing teaching into even more suboptimal methods and calling it progress.

The point of higher education isn't learning facts. If all you need is to learn skills or facts you can do it on the job, by yourself or through correspondence courses. The point is, instead, to get to know other people that also go through higher education.

Put it this way, you don't go to Harvard (or Tokyo University, or École Normale Supérieure or whichever) for the stuff you learn. You can learn much the same stuff anywhere. You go there to get plugged into the social network of graduates from that school.

That is not going to change; at least not for elite schools and the elites attending them. It's too valuable as a way to sort people into the right kind of social circles. If you get widespread non-college based higher education, you'll have it only for the disadvantaged lower classes. Children of the elites will continue to congregate the same way as now.

Sometimes you go to school to learn things from people that are wiser (not more knowledgeable) then you

For instance the Robert Wilson quote below

"In the end, Fermi convinced him to take the job by promising to meet with Wilson every Friday to talk about the physics being done. In his own account, Wilson admitted, "Sure I sold out - but then everyone has his price, and mine was a few moments each week with Fermi"

Profeessors are not going anywhere, it is just the Universities that are obsolete, except as Janne points out as social networking opportunities

Facts are one thing, knowing how to use them. I think some kind of helper will always be needed. Oh, that's what a good teacher is.
Anyway schools are where the elite met. Look at the leaders who just are not that smart, but know and are known the right people

Amen

Raw Data does not equal Knowledge.

A much better overview of the problem http://delong.typepad.com/sdj/2010/11/the-fourth-online-learning-revolution.html

A recent article indicated that the present crop of "Digital Natives" (i.e. the latest young Wesley Crushers) are quite bad at using Google: http://boingboing.net/2011/08/22/digital-natives-need-help-understanding-search.html

I'd liken education to compiling a non-trivial package of code. You have to build the tool chain and get all the dependencies first. To extend the metaphor, an accredited degree is a distro which pre-loads the basic tool chain and common dependencies and provides you with access to many but not all of the dependencies and tools needed.

At best conventional education provides you with dependencies (such as algebra or calculus) that are needed later on (for physics or chemistry for example).

The two problems I experience with autodidacts is that they often have huge blind spots (arising from missing dependencies) and they don't get the "hardening" associated with conventional education (test regimens). No learning without pain, no art without the resistance of the medium.

Are there methods better than present techniques? Most certainly! I find Khan Academy to the nascent form of this. Short good quality lectures, interspersed with quizzes to make sure the student gets it, and a clear set of dependencies. Khan himself is quite good at pointing out that most students "get" things at different rates but most do catch on.

What I would add to the mix at Khan academy is multiple paths to the same end, with different metaphors and techniques being used. Some people will understand Salman Khan,some Jerri Elsworth, and some Richard Feynman, Once this has been achieved I'd mix in a recommendation system which would identify the modalities of a student and route her on the best chain of lectures.

The other thing I'd love to see is an explicit realisation that testing is measurement in the presence of a lot of noise and a degree of Heisenberg uncertainty. If we put the onus on the tester to prove in the presence of noise what the student knows or doesn't know educational metrics would be better. When we poll or do market research on people the aim is to accurately determine their mental state without antagonising them. In essence, google-style analysing the data from the proposed uber-Khan academy to bring up the next best lecture instead of the best ad.

"The human species needs to develop ways to learn conceptual and analytic thinking other than in a classroom with other human beings."

Why?

You say that as if it is obvious but it isn't. I could say "the human species needs to develop ways to consume science fiction without the intermediation of an author." Well, ok, maybe. but why?

We have substantial scientific evidence that dialog and collaboration are essential elements in developing understanding (see Bransford, Brown and Cocking, How People Learn, which you can read for free on the National Academies Press web site).

So you're asserting a mandatory need to do something that is contrary to what we understand about how people learn and I have to ask: why? Why do you think that is something that has to be done? I don't see it and you haven't explained it.

Except We were looking ofr stable long-term institutions that work "The Unoversities" are the best working example we've go .... Erm, errr ......

You would need to keep facts in your RAM to be able to associate between them. If you have limited and narrow dataset available, then it would hurt your creative abilities.

I absolutely agree. Post-secondary-school isn't about learning information. It's about learning how to make professional contacts, work with people you don't like, and stay up all night putting together a presentation with five other people, three of whom don't want to be there. I don't think demand for those skills will ever go away.

Unless you can come up with a computer that can analyze a sophistocated argument and detect its subtle flaws, there will always be a place for professors and universities. It needs to be able to do this for arguments about things that people up until this stage haven't seriously thought about.

Can software assess the quality of a Ph.D thesis (given the original research requirement?)

The current style of tertiary education is certainly obsolete.

First, there are no 'knowledge learning' aspects one gets from a university that could not be better transmitted via lecture videos/lecture notes/readings (from a handful of the world's best professors) posted online, together with a dedicated social network, and dedicated shared workspaces in the major cities.

Second, reform the curriculum to reflect that there are actually two different types of knowledge – how to figure out which questions to ask, and how to figure out answers to questions already asked/how to complete tasks already specified – that both need to be learnt, and that each requires a different way of instruction.

Third, realise that there is no good reason for (formal) education and (formal) work to be two separate periods in a person's life. Both, in fact, should be lifelong and simultaneous. Indeed, the facet of education that prepares one for one's professional life would always, always be better, easier, and more efficiently acquired through apprenticeship for all positions (imagine a reasonably smart 15 year-old shadowing a top CEO for 7 years, and tell me that you think that at 22 he would not be able to outperform most CEOs today, let alone an undergrad from Wharton).

Universities made sense 200 years ago. When communication was slow. When the number of people interested in knowledge was limited. When compiling and accessing the findings of prior years was arduous. No surprise then that universities were effectively a bunch of smart, curious people living together around a set of labs and a major library. Today, their current structure makes no sense.

Admittedly, this leaves out one major point. The role that universities currently play in being granted access to highly asymmetrically allocated pools of capital (economic, social, political), and, therefore in social stratification. If capital is highly concentrated, then the benefits of commanding that capital are great, so many people would want it, but few people can get it. As a result, some filter system would need to be set up. As someone who did a middling undergrad degree and a top-flight graduate one, I can assure you that this filter is currently the university system. The difference in opportunities offered is stark, but for no good reason. The jobs that are filled with, say Harvard MBAs, do not actually need/use most of what is taught in such programmes. They just need a small pre-filtered pool of applicants to consider for the privileged positions. It's inefficient, it's unfair, and it also needs changing.

delurk

One obvious possibility: machine as tutor. "A Young Lady's Illustrated Primer", from Stephenson's Snow Crash, is capable of realising that its charge needs self-defence training, and providing it; Tom (Tom Baker), from Douglas Adams's "fantasy documentary" Hyperland, discusses hypermedia and Kubla Khan with Douglas Adams, and seems to be guiding him towards broadly educational goals, rather than letting him wander off on a wikiwalk, or be distracted by porn, gambling, or singing cats made out of pop tarts.

Another: outside of military war games, pilots' flight simulators, and suchlike, games and simulations are waaaaaaaay underused in education, I suspect. (See Alan Kay's "Dynabook" concept, LOGO, and http://worrydream.com/KillMath/).

However, Alan Kay has spent decades beating the Dynabook Horse, and, as atimoshenko hints, cultural and economic factors may well be in the ascendent here: universities work for signalling status because they're expensive, and producing better, cheaper, faster, out-of-control-er education using technology may be somewhat futile.

I don't think universities or professors are going anywhere. Universities appear bloated for a reason. They are fulfilling several roles in modern society, and there are several organisations for each within the university. The only one that Joan brings up is education. In addition, there is research and public information (those are the three areas formalised in Sweden).

The problem with university education is that it serves a two-sided goal. The society wants skilled workers for jobs that require academic education, while the universities (more exactly the professors) want to educate new professors. In some way the goal is even three sided, because society also wants to increase the average level of education since this has positive effects on health and welfare. Already today there are partially separate internal organisations in the universities to handle these different types of education. There is firm evidence that teachers are important, maybe some really good teachers can produce teaching material centrally, but someone also needs to supervise and grade the students. Remember that we are trying to get everyone through higher education today, not just the most motivated, this requires teachers. Whether you want them to be called professors or something else doesn't really matter.

Then there is research. There is quite good data showing that the larger the scientific environment the better the research. The biggest institutions are producing the highest impact science, and they are getting even bigger in order to be able to produce even better science. Sadly industrial basic research is declining. So, the big research institutions aren't going anywhere. Again, it doesn't matter if you want to call the research group leaders professors or something else, the role will continue to be needed.

Finally there is the informing of the public. This goes hand in hand with research. It is about educating the public and the public servants about what the newest research indicates would be good public policy. It needs to be done and forcing researchers to do some of the work themselves in order to become professors is a way that has worked (for some definition of worked) in the past. At least researchers in collaboration with journalists, teachers and lobbyists can hope to bring public opinion and policy closer to something that has backing from scientific evidence. This role needs to be tied to those involved in the research directly, at the same time it is good if they have some experience and authority. The very hard vetting before becoming a professor at a top institution and the constant peer-review and public scrutiny of what they produce serves to safe-guard the quality information that is used to form public policy. Although, it is by no means a guarantee that the advice will be followed or that it will work.

Together, I think there are good arguments for maintaining the universities and having subject specialists, who both teach and are involved in original research, with an explicit mission to inform the public about recent breakthroughs and their implications for public policy. You may call them professors, or something else, it doesn't really matter.

If memorization is what you're doing in University, you're doing it wrong. The point of advanced education is understanding, not memorization. Classrooms, maybe, are less important than they used to be. (I am not sure they were ever a good idea.) Self-directed learning seems to be becoming more important.

Google replaces memorisation. Most universities don't realise that yet, so yes, hell will break forth when the realisation finally hits. Over priced and under performing - the certificate can be replaced.

As for what are professors for?

University departments: "a more retched hive of scum and villainy". University departments are to keep those who would become master evil protagonists from being unleashed on the world at large. Just think what would happen if that brainpower - that currently goes to snipe at colleagues - would do if turned outward.

There is a huge difference between knowing and understanding. In the short term brain computer interfaces will let you access (say) the proof of the 4 color theorem in minute detail - but you will still not understand it.

@Atimoshenko

So, if you wanted to learn how to do leading-edge scientific research, you would advocate that people should work in an apprenticeship with leading-edge researchers? This is what research-active universities provide for their postgraduate students.

Ditto for training medics, and presumably for at least some other subjects too.

Now, whether lectures still have a place in tertiary education learning is another matter. At our uni, we record many lectures so that students can revisit them later. Yet most students still choose to attend the original lectures as well.

If memorization is what you're doing in University, you're doing it wrong.

This.

For that matter, the same probably applies to all education... There's just a set of perverse incentives that tends to gradually turn a useful, important goal like "familiarity with the periodic table" into soul-sucking drudgery of the "memorise the first 20 elements" kind — not only entirely useless, but also betrays that even the person setting the curriculum has failed at the original goal, let alone the teachers and students.

'The bloated university structure is--literally--medieval.'

The bloated bit isn't the medieval bit: it's the partially-successful takeover of an efficient and egalitarian medieval institution by the structures and techniques of large, hierarchical corporations.

'all facts, knowledge, and experiences are retrievable instantly without mediation'

The things most worth knowing and experiencing are difficult, and the best way we yet know of to help people get into them is being taught by people further along the path. Like, say, university professors who've spent their lives working with that knowledge and experience. I mark enough bad student essays to know that having instant access to a summary of Hume's philosophy is not equivalent to understanding it.

I agree with the statements about university being about making contacts, learning personal management skills (applying them to yourself and others) with the actual "content" being secondary.

Once you've proven that you have some sort of academic proficiency/intellect and a desire to pursue this, self-management skills and contacts become what elevate you, as you would be able to excel at Astrophysics OR Burkina Faso studies.

What I would like to add to the mix is that the feeling of being forced to learn things "slow" frees the idle intellect to make it's own connections and thus spawning new ways of thinking instead of being "forcefed" pre-digested knowledge.

I believe that most profound insights come from being immersed in existing knowledge (+ some degree of idleness).

If an actual learning process has to happen, beyond a rote recitation of facts, the brain needs time to make the connections on its own. That's when being at a classical university helps, because you can immerse yourself.

If we were to synthesize this process in silicon or chemically boost the biologicals at hand, then this of course becomes a whole different story and I can see 100% online/memory packaged educations coming to life.

But I think that because of the way we absorb knowledge in order to actually learn and make working connections in our brain and with other people, the "campus" strategy is not going to go away soon.

Whether it'll be in Cambridge, LEO, Mars or virtual space is another matter.

What is the huge difference between knowing and understanding ? We use these everyday terms and obviously perceive a difference between them. But neurological how do they actually differ?

The difference between knowing and understanding is the difference between reading Wikipedia on neurology and being an actual neurologist. And the Cathedral isn't going anywhere soon.

Some universities are rather adept at adapting to the times, and imho there will will always be entities calling themselves universities. Some will be research centres, especially where there's a lot of heavy lifting and no benefit to decentralisation. Some (in the UK) of the newer universities may go to the wall as propective stuents do a rational cost/benefit analysis. Others may become simple holiday camps, or even living history re-enactments for tourists, convention centres, "thinktanks" of spurious utility etc.

I should expect All Souls (Oxford) to continue in its present form for a very long time.

I do not know how knowing and understanding differ, but when I am learning something new and difficult it is very obvious that there is a difference that can be felt. Quite often I read the words or equations over and over again, then at some point, usually after sleep, it suddenly seems "obvious", whereas before I might follow the details but not grasp the big picture. I would guess that the process involves a conversion and integration of the rote learning into the existing knowledge structures in the brain.

The key is not knowing, it is understanding. Thus you learn to seek knowledge and to connect it together in a structured way. The fact that you will learn some facts along the way is purely incidental.

The best education is when instead of knowing the formulae and constants for 100 different physical effects and laws, you'd rather know of existence of 1000 different effects and laws and could understand when which of them applies and look up relevant formulae and constants in a couple minutes.

But can you then build understanding without first acquiring knowledge? A trained neurologist would have 2 different sets of skills. There is probably some relatively automated procedures, specialist skill sets. Then there is the broad understandings, associations between different knowledge.

But both require you to have acquired facts about the world.

@pjcamp, let's try out some arguments to back up this assertion:

"The human species needs to develop ways to learn conceptual and analytic thinking other than in a classroom with other human beings."

The global economy is only exacerbating a trend that looks like a long term shake-out in higher education--the weak schools get weaker and sometimes close. The elite schools get stronger. The schools in the middle will either slide back toward extinction or find their niche in the elite.

In the post-oil world, moving human beings is going to get more expensive. To get to an elite school with your specialty you may have to travel halfway around the world. Consider the energy costs of moving the protagonist in Joan's book from her home to school in the space habitat. Her family had to be very wealthy. The political and epidemiological arguments are one's you've heard before or can construct yourself.

People are doing some tremendously interesting work to improve online education and perhaps we can increase the bandwidth of sensory data enough to mimic reality. Even so, odd things like time zones get in the way. My son and his Serbian chess teacher have both made extraordinary efforts to overcome a mere six hour time difference. Anecdotally, asynchronous online learning seems to be more successful than synchronous online learning. At least in Western culture, we can't even be bothered to watch television shows all at the same time anymore--except possibly sports--and Tivo time shifting is the dominant meme. For these reasons, I'm skeptical about the replacement of the physical classroom with the virtual one.

Memorized knowledge is good for solving utterly standard problems. Understanding is required to adapt that knowledge to novel situations.

Scifi referent: The Bobble Wars

Professors are vital, the academy is dead. Learning analytic thinking is more important than ever.

This is, perhaps, U.S. Centric, but there is an endless stream of editorial bemoaning the numbers of English majors, instead of engineers. Most of those, if you look, seem to be written by people with law degrees.

Perhaps we are getting even more medevil - training via bloodline, or mentor/guild admission.

Agree. But that also mean that you can never acquired understanding without first memorizing knowledge. As my sensei would tell me first you repeat until you can imitate perfectly, then you start improvising.

That's not necessarily true. If you are good, you can extract "understanding" from a much smaller body of memorized knowledge than other people. Do we have to memorize a list of everything that falls when you drop it, or just understand the concept of gravity?

In one sense, knowledge is data and understanding is the data compression algorithm.

As above, lectures, etc. are obsolete , but professors are vital. (I disagree about Academies being obsolete, but their role is fluid, as has been pointed out).

Lectures serve two purposes, often confused: (1) dissemination of information, now made mostly obsolete by the Internet. You can get the information easily, in most cases. (2) A course in what information is important, and why. This is simply not known to the student beforehand. Even when the lecture is given electronically via podcast, etc. this bit holds its value.

As a software engineer I am long accustomed to learning the basic ideas, and lookup up the details as needed. Similar to the offline memory concept. However, when I learned to fly, I realized a different style of learning was necessary. I couldn't pause things to look up some detail.

While software development and flying themselves may not be directly transferrable into the future they are representative types of skills that will remain.

Sometimes you can rely on offline storage. But sometimes you've just got to really know it!

If I understand you correct, you are stating that understanding comes from building models based upon perceived reality. But is it possible, or efficient to learn such models directly, without having an established context of knowledge that makes the model relevant. Or to turn it upside down. If there should be any use in compressing data, then you first need to have a full harddisk.

The data compression algorithm (rules) can be extracted from a very tiny subset of the actual number of possible examples.

Then there is the meta-knowledge (or understanding) of how to apply those rules.

And before those rules can themselves be converted from data to knowledge there has to be a mapping of structures eg to internal brain states which integrate them with other sets of rules etc.

Of course, we can and are taught the rules as well as the examples. Otherwise every student would have to reinvent all of mathematics and science for themselves.

As, for example, you can either learn the various equations for Newtonian mechanics, or you can just learn calculus.

You might learn about gravity as a model in physics class, your paper calculations might get an A. But you will not truly understand gravity before you fall down from an apple tree.

What is the huge difference between knowing and understanding ? I think Aldous Huxley explained the practical difference well in Brave New World.

Knowing is learning the fact "The capital city of France is Paris."

Understanding is being able to reverse-engineer the learnt fact to answer the question "Paris is the capital city of which country?" by saying "France".

Knowing vs. understanding: I've had any number of lecturers (professorial and others) who lecture about a theory, required us students to memorize a formula, then expect us to be able to apply it on a test.

Obviously, knowledge doesn't work that way. Today, we've got kids in college* who couldn't hand graph out y=X^2 or y=log(10)X. They have no sense what those curves look like, what they mean. This is a problem, because it means they don't have an intuitive grasp of non-linear growth, such as the interest on their student loans, population growth, and so on.

This came from an ecology professor who started handing his students in general ecology a pop quiz in math the first day. Most of these college students naturally operated at a pre-algebra math level. He instituted remedial math in his general ecology class, just so the students wouldn't get totally lost. The sick part was that the students I talked to in this class were defensive about forgetting, saying it was normal for *any class to forget everything after they got their grades.

Universities are not going away because for some fields, like the sciences, you need people who are doing research and the facilities to support that. No amount of online learning will give you hands on access to expensive equipment and direct instruction in its use.

I also believe that the small group learning model works very well in improving learning.

What should disappear is the attempt at scaling by having large lecture theaters. Despite the attempts to improve the quality with various technical aids, this still seems to be a mistake that could be replaced with a good canned lecture watched remotely.

Universities in California (and I assume elsewhere) seem to be bloated by huge administrations. Some administration is necessary, but the support staff vastly outnumber the teaching staff. This looks like an organization problem/failure.

Britain has the "Open University" that works on a different model. Does anyone know how the quality of that education compares? I remember it being looked down on, but what have studies shown as to it's effectiveness?

Memoization is both a cognitive load and a benefit for thought. Books (online access) clearly can reduce the load, but it is needed to build understanding by creating new connections in the brain.

I think the failure of an "Encyclopedia Galactica" is an imagination failure. It is not unlike Searle's Chinese Room thought experiment, or the difficulty of understanding how neural wetware can lead to the qualitative state of minds. We simply cannot literally see how our eyeball access would give the user teh bandwidth to use it like an extended mind.

However, if we envisage a global brain that mimics human brains and in which you could immerse yourself, so that it felt like you had a much larger brain, then I could see that learning might become very different.

The professor before an audience of a thousand has long been obsolete. Back in 1999, professors at UCLA invented something called Calibrated Peer Review that replaces the professor with a computer plus student interaction. The students without the professor got better grades in organic chemistry than they did in the traditional class with the award-winning professor who invented the program.

Medical diagnosis is also quietly getting replaced by computer programs. Not surprising--in the old days, medical students were taught to hold the Physician's Desk Reference below the level of the patient's bed so the patient couldn't see they were looking something up.

What can't yet get replaced is the one-on-one exchange that goes on in small colleges, between the brilliant academic and the unprejudiced novice who doesn't yet know the dogmas. The kind of games Socrates used to play with his cute students. By the time machines learn to do that, they'll also have gained the emotional intelligence to demand a salary and retirement; the kind of annoying demands that society wishes not to pay taxes for.

I find lectures by people (Profs, graduate students, random weirdos) essential because they are given in a different format in a different way (by a person using their voice, often with some slides and drawings) which backs up the textbooks or papers which have been read before and after the lecture. And you get to ask questions, which is usually very useful for elucidating points you don't quite understand. What computer program or recorded lecture can do that?

"And you get to ask questions..."

Which is essential. Or rather, getting the answers is. This is where, as an engineer, I find user communities like Stackoverflow to be of immense value.

We teach people to put apostrophes in the right place. I have an interest to declare about the OU*, but I think that it works as well as any institution can which is educating adults part-time. By the way, we tried pointing a camera at a lecture and discovered that this doesn't work very well. Making a dedicated TV programme instead works, but is expensive: too expensive for us to do any more.

By a curious coincidence, last week I was back at my old college in Oxford (taking the opportunity to tell them that the Open University is the best university in the world...), talking about the use of the tutorial system in my education. It didn't help me very much up to the point of getting my doctorate, but it's done a lot since, notably with the 'public outreach' function of the university mentioned above. The trouble is, it doesn't half cost . . . About sixteen grand a year. Perhaps you could cut that cost down to about twelve, but not a lot more. Meanwhile, everyone else can do it on about seven grand, and it costs us five. We win. Again.

*It pays my wages.

"Making a dedicated TV programme instead works, but is expensive: too expensive for us to do any more. "

Now I'm not clear about the relative costs. It is too expensive to make a program (those old ones with profs in 1970's wing lapels still being used?); so how does that equate as the total costs being lower than traditional universities? Does the scaling inherent in broadcast not translate to savings when other administration tasks are included?

Fess up - University (without a stunning reputation) lecturer 1. The education you get at university has nothing to do with the course you are studying it has everything to do with the opportunity to meet with people who have the time and inclination to question anything and everything about their existence and given wisdom. 2. If the education you get from a university is good you learn to question everything for the the rest of your life without becoming paranoid or paralysed with doubt. 3. An ideal university education gives you the curiosity of a three year old with the pragmatic reasoning of an Agatha Christie detective. Humans are primates - typical pack animals who employ distributed intelligence through job specialisation - a key specialisation is knowledge developer and accepted knowledge questioner - there is a reason these are the first pack members to be shot after a successful revolution - they question everything. Ok that's a very modernist view - but hey isn't neo-modernism the new cutting edge.

My understanding (from an OU lecturer other than Chris) is that academic educational programming costs tend to converge with typical high-end documentary program-making costs, i.e. the sort of stuff you see on the Discovery Channel or Natural Geographic, only with more substance and less fluff ... around US $1M/hour. Which is cheap by modern TV standards -- anime and/or cartoons cost about that much to make; top end TV drama from the major US networks are around $10M/hour, while BBC drama programming is somewhere in-between (i.e. $0.7M/hour - $3M/hour) -- but prohibitively expensive if you're a university teaching 20 different subjects to degree level with maybe 150 hours of lectures per student per academic year. Especially as the content may have to change from year to year to keep up to date with progress is the field.

So excluding admin costs, that means ~$200MM/yr production costs.

If students pay $10k/yr fees, the breakeven = 20k students/course.

That seems about the ballpark size of the UC system. If content doesn't need to change every year, then the breakeven student body/course content falls rapidly.

Conversely, to lecture those courses (excluding labs) with 250 student auditoriums requires 2.5 full professors (30 hrs/semester)x 20,000/250 x $100k each = $20MM/yr.

If the class sizes were closer to 25 (tutorial/discussion group sized) the costs would be comparable.

Of course there is overhead - a lot of it in buildings that is not needed for broadcast.

The OU broadcasting approach should scale better assuming that testing could be mostly automated.

Perhaps the OU is the better model for nations without an established university infrastructure and large populations?

Not done a vocational or professional degree, have you? I'm betting you're coming at this from a liberal arts perspective. Am I right?

"So let's postulate a near future like the one in Accelerando, where the brain-Internet interface is much better. Why do we need content memorization? Or is it important to have some data in RAM, so to speak, and not just available on the hard drive?

There's a decent chance that yes, you do. Critical thinking skills are not cross domain, that is, a physicist can be brilliant at being a physicist, but he won't have a much better chance at making the right decision on pursuing a given alternative medicine therapy. One of the theories on why this happens is that you have to have a large base of related knowledge to work from in order to apply critical thinking skills.

It's possible there's some other reason of course, but probably expertise is still going to rely on having a ton of knowledge in your head, even if the specific knowledge for the problem you're working on mostly comes out of Google Scholar.

The real change might be in terms of continuing education, there are certain fields that can change blindingly fast, or even at a moderate pace, and I've seen professionals who still do things the 70s way, and been caught myself on things that were called into question within a few years of the college course I took. Technology may go a really really long way in terms of making sure the info people use is up to date.

Interesting question. I don't think professors are obsolete, but I think the skill sets may change to look more like Teresa Nielsen Hayden's than a traditional lecturer as more courses move online. The reason I say this is that the best classes I took had some sense of community than not. When I moved to online classwork, I found that sense of belonging and community gone. And that the professor hadn't the faintest clue of how to encourage it.

Somewhat related: Anyone have a course on how to evaluate the sources and biases of an instructor/reader/presenter? I think learning how to assess those might be more valuable than anything else. Could they be taught at a university? Maybe, but I suspect most professors would be scared of the techniques being aimed at them.

Also somewhat related: Since so much of how classes and learning are evaluated are tied to rote memorization, why are not more folks teaching memory tricks to kids? From peg words to Dominic's hotel, to mind maps, to silly songs, etc., you'd think there would be a definite interest in getting the most bang for buck per unit time involved instead of grinding memorization. Anyone able to comment on this?

Universities (and schools for that matter) will stay important for the transfer of information that is never written down or well documented. Especially in the sciences, but I imagine also in other fields, there is much that is taught by example, direct instruction and most importantly error correction.

Since so much of how classes and learning are evaluated are tied to rote memorization, why are not more folks teaching memory tricks to kids?

We like to think that children learn more than one thing in school. Teaching kids memory tricks would involve admitting that there is in fact only the one. Most people like to fool themselves, especially when it's the social norm.

Those who do admit that school typically only teaches one thing instead tend to look for a school that does better, some form of alternative schooling or (in extremis) home schooling. Few will consciously choose for their child to spend 10-20k hours becoming expert at rote memorisation.

@pjcamp, let's try out some arguments to back up this assertion:

"The human species needs to develop ways to learn conceptual and analytic thinking other than in a classroom with other human beings."

The global economy is only exacerbating a trend that looks like a shake-out in higher education--the weak schools get weaker and sometimes close. The elite schools get stronger. The schools in the middle will either slide back toward extinction or find their niche in the elite.

In the post-oil world, moving human beings is going to get more expensive. To get to an elite school with your specialty you may have to travel halfway around the world to get in the right classroom with the right professor. Consider the energy costs of moving the protagonist in Joan's book from her home to school in the space habitat. Her family had to be very wealthy.

People are doing some tremendously interesting work to improve online education and perhaps we can increase the bandwidth of sensory data enough to mimic reality. Even so, odd things like time zones get in the way. My son and his Serbian chess teacher have both made extraordinary efforts to overcome a mere six hour time difference. Anecdotally, asynchronous online learning seems to be more successful than synchronous online learning. At least in Western culture, we can't even be bothered to watch television shows all at the same time anymore--except possibly sports--and Tivo time shifting is the dominant meme.

These are just a few musings which came to me as a part of the process of student-dom.

When I first headed to university straight out of high school (all those years ago, back when we rode dinosaurs to campus) the difference I found was this: in high school, we were given a treasure map, with an X marking the spot where we'd find some buried bones. Follow the map, go to the spot marked X, dig up the bones, bring them back, and you got your passing mark. University was different. The university just had a load of pieces of paper, sitting there. You could pick them up - or not - at your own leisure, and then you could fill in your own map, and dig wherever interested you. Maybe you'd find buried bones. Maybe you'd find nothing. Maybe you'd hit a water main. But maybe, just maybe, you'd find some buried treasure.

Okay, it's a bit of an elaborate metaphor, but that's what the difference between high school education and university education felt like to me, as a shiny new university student back in 1989. Since then, I've been to quite a few different tertiary institutions, and started a number of courses, and the best ones, to me, have been the places where I got that "the pieces of blank paper are over there, grab one if you'd like" feel about them - they're the ones which encouraged me to think a lot more.

As someone who arrived absolutely and utterly jaded with the school education system (and who would have gone into hard revolt if faced with another three years of same, no matter how damn rewarding it was in the long term) it was a refreshing look at the differences between education and learning.

That's a shame. Because if they learn the tricks, teachers and students could spend less time with the mind numbing drudgery of memorization and dedicate the rest of the time to learning. Though I suspect that would only last until administration caught up with it...

Granted, I'm about to teach peg words and some other tricks to a friend's kids. We'll see how well this works and whether or not they get bored.

In answer to your question:

"People working in education have always wrestled with the "skills vs. content" problem. But the way we make connections and approach subject matter has changed since the time of Aristotle. The human species needs to develop ways to learn conceptual and analytic thinking other than in a classroom with other human beings. Does science fiction offer plausible alternatives?"

Different countries offer a lot of alternatives. The past does too.

I learned all of my conceptual and analytic thinking a long time before university because the private schools I went to started at elementary level (none of them teaching in English)to emphasize Cartesian logic. They did so even before the precise split-up years when we were put in concentrations leading to our future, destined academic careers, in engineering, law or medicine. This meant an intro to logical thinking in our early teens and chemical and physics labs not long after. It was as if prep schools started right after we could read.

Before my time, before schools became widespread with the introduction of cheap paper based printing and steam presses people with the means to buy those extremely costly books also employed tutors and governesses. Often their children would be introduced to conceptual and analytic thinking and to hands-on lab work even earlier than what I and my former colleagues experienced.

I don't think you need what's called "strong AI" to mimic precise aspects of the work of a tutor, then other precise aspects of the work of a governess, and precise aspects of the work of a music teacher, and so on. All you need are smart parents who can plan those interventions and balance them with social activities like horse riding or Olympic gymnastics.

So, I think that most universities/colleges in the US, and the English ones in Canada, will eventually shrivel and die because children will have learned conceptual and analytic thinking about 10 years earlier than the current practice.

Things will be different in the UK as universities will barrel on with the sheer inertia of centuries of tradition.

A big part of learning to do things is practice. It helps a lot of be around people learning the same thing, especially when what you're doing is hard. I found ordinary differential equations quite tricky to learn and never mastered them. Hanging around with the smart kids helped me learn. And ODE's are the basis of a huge amount of electrical engineering (and I suspect other sorts too).

A muppet who can plug values for an ODE into their prosthetic brain and find a solution is useful. But they're going to be doing a lot of typing and waiting. The geek who can look at an ODE and have a pretty good stab at a solution in their head, then optimise that solution without recourse to a machine will have finished the job and left the building while the muppet is still wondering if there's a pattern there somewhere. On top of that basic skill you pile a bunch of transformations that can (sometimes) map intractable DEs into tractable ones. At that point your geek is starting to become useful and may eventually be able to (say) design an aerial or modify an amplifier design to make it stable.

A lot of what universities do is that: put people together and get them used to solving problems with what they have to hand. Plus teach them to go out and add to their pile of "what is to hand". The difference between a novice and an expert is a lot to do with the number of potted solutions they have in their head.

Oh, and one reason for formal schooling/undergrad degrees is so that those really good professors don't spend their days giving basic remedial education, hopefully they too are learning by teaching. For all that it's kinda fun to take image processing lectures right back to Maxwells equations, that's not really a good use of everyone's time after the first time. But if that professor was instead teaching us our times tables or basic spelling and grammar... that's just dumb.

"Google replaces memorisation"? I wish you luck learning math, a language, or a body of literature (history, a law code, a corpus of literature) without memorization. Memorization is a fundamental part of the learning process; a significant part of the curriculum of medieval universities was learning how to memorize efficiently. Unfortunately, our culture devalues it; in part because we are sometimes expected to memorize the wrong things.

There are different ways to memorize things, but if you can list ten databases and websites that are most useful in your field then you memorized them somehow. Writing and search shift the balance from learning things to learning how to find them, but every piece of information you have to look up is a distraction from a harder problem ("what does this word mean" vs "what does this sentence mean"; "what is that part" vs "why does this engine keep catching on fire").

Actually Charlie my degrees are in physics and computer science although my research area is socio-technical design for information management. I spent a lot of years developing systems in the real world before entering academia so yes quite vocational and professional and quite certain education is about developing enquiring minds not about training them in the delivery of shrink wrapped solutions.

Charlie oversells his "education bubble" concept. Ockham's razor tells us that if there is high graduate unemployment, a simple explanation is that there might be a recession on! And if we look out of the window...why yes.

Further, student debt is a thing in the US and to a lesser extent the UK. Not, say, in Germany or China. Are you thinking what I'm thinking?

The obvious answer to (some) student debt is for the govt to give full grants to (say) the top 5% of students and/or those doing studies that will directly benefit the nation eg science rather than media studies.

The rest can pay for themselves if they think it is worth it.

"One memory trick"... not really.

I spent some time as an instructor, planning and conducting training for (part-time) soldiers and officers.

One interesting article that I came across was that parable-based explanation was twice as likely to "stick" in someone's mind as a recitation of facts. Rather than saying "don't do X, it's risky", you give real-world examples of people who did X, and suffered for it. I'll vouch for it - it works.

Part-task simulation is also very useful; you work people up gradually, before you try and exceed "full rate" simulation. Train hard, fight easy is the oft-heard cry - "Kobayashi Maru" is not a surprise to anyone who has been through a uniformed services training establishment.

Instructors also have to structure the teaching of the subject matter so that the students are making small steps, not huge leaps - and are able to relate it to existing experience. For instance - the British example for teaching a structure for "how to solve a tactical problem" tried to relate it to a real-world situation to illustrate the steps involved; and the examples in the training matter chose "how to buy a car". This is great if you got a bunch of twenty-somethings who've actually owned a car. Unfortunately, I was teaching a bunch of non-car-owning nineteen-year-old students, so I rewrote the example as "how to organize a party in your flat".

A minor point about the original post... I agree about the disappointment, but:

Star Fleet academy is both a little disappointing and a bit reassuring; it doesn't seem much different from current military service academies (the Kobayashi Maru simulation notwithstanding.

... from current US military service academies.

Other countries do things differently. West Point and Annapolis are slightly unusual Universities - the students spend several years there wearing uniform, they spend most of the time in the classroom, and they leave with a degree (and a very chunky ring).

Sandhurst, Dartmouth, and Lympstone (the nearest British equivalents to the above) are more vocational training establishments; the training lasts no more than a year, and while there is a theoretical component, the bulk of the course is as practically-based as possible. About 80% of the students are post-graduates.

Did you mean "there are students who can't hand-sketch an exponential growth curve"? At first I read your statement as "there are students who can't plot co-ordinates on graph paper, and then draw the best fit curve through them".

On that general subject I can't hand-do calculus any more; I've never used it since leaving full-time FE. OTOH I can do algebra and geometry because I use them most days.

"Google replaces memory."

Not entirely. Even when all the data is available via Google, you still need a memorized base of knowledge to be able to find what you want.

My friends often ask me to identify plants, after they have spent time fruitlessly searching on google. I am not a botanist, but since I am interested, I have memorized the appearance of a lot of plants. Consequently, I can often identify the Family, if not the Genus, of a plant by glancing at it. Knowing the Family and the place where the plant was found, it is often possible to identify the species in a few minutes via google.

If you haven't trained your brain to notice and recognize details of flower structure, then the flower will be be a colorful but vague "thing" and even the fastest connection to Google isn't going to help.

The natural sciences are all like that. Before you can analyze, you first have to train yourself to see. And seeing, really seeing, requires memorization.

"And seeing, really seeing, requires memorization."

And that is a very difficult AI problem once you get beyond toy problems.

I agree that Google doesn't replace memory. If you don't have an understanding of the subject, then you can't tell whether what Google brings up is reliable information or nonsense, and there are topics where nonsense dominates the results.

What Google does replace is a library. Unless you're working in a major research university, Google gives you access to information is greater breadth and depth, and more conveniently. The problem remains that much information is paywalled, at exorbitant rates, or is not available at all online.

For an example, a couple of years ago I wanted to refer to an 18th century botanical work to clarify a nomenclatural point. At the time it wasn't available online, but it wasn't available offline either (I tried the Royal Horticultural Society library). No doubt I could have got my hands on a copy by interlibrary loan, but that would have been a lot of hassle to read one page. The work is now online, and the question answered.

Have you seen google goggles? Point your smartphone at a thing and it will tell you what it is. I don't have a data plan so my experiments have been limited to household items, but I'm guessing it wouldn't do too badly with plant recognition.

My latest botanical google search was to identify the strange tree with the stinky fruit in the park as a ginkgo, I'm certain goggles would be able to match the distinctive leaf shape with an image search.

There's been a bit of to-ing and fro-ing upthread about the relative differences between professional/vocational qualifications and humanities/social sciences ones. For what it's worth, I suspect that, in the medium term, automated hypothesis-generating algorithms may well change the face of scientific education beyond all recognition. It's already the case that elementary physical laws have been 'discovered' by automated algorithms. Extrapolating this a bit further, I wonder if we're going to come to a point where scientists don't even engage in the game of hypothesis testing, and merely chase the results that are presented to them. And if the structure of the universe is too complex for the human mind to grasp, well, we're in a wold of hurt then.

Ironically, I can't see this happening (or at least not so fast) with the humanities/social sciences. No algorithm I'm aware of is yet capable of, say, performing a convincing semantic analysis of a novel; and any ones projected in the past have been dismal failures. (I should know: my PhD involved trying to develop one.) Whether this is because no algorithm exists for semantic analysis, or that one exists and is fiendishly complicated is impossible to say. What is certain, however, is that any discipline that involves internalising algorithms is likely to change dramatically. After all, my mobile phone can already do Sodoku orders of magnitude faster than I'll ever be able to, but it still can't get predictive texting right. We may all end up in the arts faculty yet ...

Universities aren't going anywhere. They will change and ironically go back to being even more medival. They will become gathering places for intellectuals to follow their interests, aka universities will go back to being mostly research facilities with a teaching sideline.

Google will not replace a good teacher but the internet has changed how things are being taught. Most (but not all) of what is taught at the undergrad can be done online and in a simulator. Reducing costs by reducing the need for faculty and facilities budgets being wasted on average students.

It also givse the university the ability to screen candidates for expensive face time with proffs. This screening reduces risks to the the student, the university and society education ain't free regardless of system. Thats money, facilities and man hours free'd up for research.

As for universities being a place to make contacts, bupkis. It happens in a few occasions but most people are too young,lack the judgement and haven't earn enough trust to make proper contacts. By far the best places to make contacts are in Trade Associations, internships and on the job.

The best practical education I ever recieved in making contacts and people skills was in the military. Watching a couple of E6's and E7's working a "drug deal" (figure of speach not an actual drug deal) was a site to behold. It started with a game of 6 degrees of seperation to make a common bond, and once that was done they got down to brass tacks, and the deal got made. Millions of dollars of equipment could be "loaned" (all hand receipted of course) to other units in exchange for favors held in reserve, usually in form being "loaned" other goverment equipment you are short on, soemtimes even extra personnel.

Don't get me started on Sargent Majors E9's the mafia dons of the military. You start learning those people skills in the junior enlisted ranks. Its how you get promoted

Speaking as a botanist, um, ginkgo's one of the easiest plants in the world to identify. No other tree has leaves like that.

We'd love to have google goggles (think of how much easier conservation would be!), but I don't think it's going to replace botanists any time soon. We already have substantial online resources, and believe me, we use them. The problem is that the critical details vary among species. In some species, smell matters. In others, tiny hairs are critical, or details of leaf shape. In wild onions, one key depends on the scale shapes on the bulbs underground.

And some species are simply hard to see, because you need to develop a search image to find them, or they're tiny. This can be pedantic, but a number of endangered species are actually a pain in the ass to find and identify.

Also, it's often easier to simply get trained. A half-hour with an expert and a plant species will let you identify it with confidence for years, but working from an online source or book can take days and result in confusion.

Technology helps, but it won't replace humans any time soon.

The problem with modern universities is that they have allowed vast numbers of unqualified students into their ranks for the sake of various leftist political ideals, and consequently have greatly dumbed-down and politicized the academic environment. Spend some time at a secular university in the United States today; what you will find has as much resemblance to a Marxist indoctrination camp as a place of higher learning. I will therefore be quite happy to see these increasingly useless and Soviet-style bureaucracies receive the gutting they so richly deserve.

Plants are very challenging subjects; they often look very different than their "photographs." They grow in different forms depending on their environment, and they can reproduce asexually or not. My lab assistant raises carnivorous plants, which have tricky requirements including distilled water (how they manage in nature, I'd like to know.) He recently left for plant grad school and is looking at Martyniaceae. BTW Heteromeles is a beautiful low-water shrub.

Most carnivorous plants grow in bogs and swamps, and tend to grow on top of / in moss. They don't need distilled water so much as they need water without certain common salts and minerals; distilled water is just the easiest way to mach what they get.

This is also one of the reasons most of them are dying in the wild -- contaminated water.

I think you've stumbled onto a secondary survival strategy for carnivorous plants, by the way. Some creatures get along by being so useful and/or weird that humans will pamper them, and so reproduce despite being unable to get by in the wild.

Speaking as a botanist, um, ginkgo's one of the easiest plants in the world to identify. No other tree has leaves like that.

Speaking as a non botanist, I know that now, but two weeks ago a gun to my head wouldn't have helped.

I identified it from the stinky fruits through an old fashioned google text search, and the distinctive leaves is why I'd rate goggles' chances - the tool isn't miraculous just yet, it mostly matches images with other identical images and will cheat if it can (Like reading a barcode to identify the product, or simply find the exact photograph it's seeing in a reverse search rather than identify the contents)

Still, it's rather a game changer to have that kind of potential in everyone's pocket. The app is practically new after all.

What happened to that mechanical marvel of yesteryear, the mechanical educator?

That used to be pretty big in sf; you had Seaton,Crane and Duquesne using one to instantly learn an alien language and then later to acquire an intimate and encyclopediac knowledge of various arcane sciences. Later novels based their economy around them.

If anything, this is one area where sf has taken a giant step backwards.

There's an app for that - "Leafsnap" on the iPhone.

It is very limited and I don't think it works well, but it is a start.

Just so everyone knows where my interests are vested - I teach math at the college level: algebra, calculus, statistics, etc.

And to those people who think that you can learn this stuff on your own through some sort of computer guidance, well, I say (with a certain amount of experience others seem to be lacking) that's a load of hooey. I've given tests where students are allowed one two-sided page of notes where they can copy down all the integral tables they could possibly need for one of my exams - and most of them still make a pretty poor job of it.

As other people have alluded to already, there's a lot more going on than rote memorization. Something that can't be encoded into a decent reactive teaching program at this point in time, at any rate.

I recommend this article by Adam Gopnik in the New Yorker, about machine intelligence. http://www.newyorker.com/arts/critics/books/2011/04/04/110404crbo_books_gopnik

Gopnik argues: "The Singularity is not on its way—the Singularity happened long ago. We have been outsourcing our intelligence to machines for centuries. Now they are much quicker at calculation and infinitely more adept at memory than we have ever been. And so now we decide that memory and calculation are not really part of mind. We place the communicative element of language above the propositional and argumentative element, not because it matters more but because it’s all that’s left to us."

And speaking of steps back... The End of the Future, by Peter Thiel: http://www.nationalreview.com/articles/278758/end-future-peter-thiel?page=1

These attempts to automatically generate fundamental equations crop up quite often and they are not without merit but they typically only work on well understood domains because they are carefully constructed to find what is known to be there. This program appears to be a random search with a fitness function and that's where the problem is - the fitness function has been given having been worked out from the already understood domain. So it is basically a long winded way of finding close approximations to things that are already known precisely. Rigid body motions is an important area of study in theoretical robotics but researchers are more likely to use thing like Lie or Clifford Algebra's which are capable of characterising a whole range of motions in a very succinct form. Of particular interest are singularities :) - which you generally want your robot appendages to avoid. What is interesting form the POV of this discussion is that there is a huge body of work completed by mathematicians in the 19th and early 20th Century but it is largely written in German, French and Russian use now obscure notation and technical language that does not match the way modern researcher pose their questions. I bit of AI that could overcome those problems would be immensely useful. Probably on a par with semantic analysis of a novel though :(

I've seen a couple of papers on computer recognition of plant leaves. It wasn't very reliable.

Teachbot http://www.teachbot.net/ from Elzware is pretty cool and has been very effective in the classroom. It has two or three features that I think are very important in educational software. It practices unconditional positive regard - it doesn't care how many time you ask the same question, it provide internal locus of control to the student and to the teacher - the former can direct their own learning going where they want to go, the later can see how students are doing individually and collectively and plan interventions for any areas where people are having difficulties.

Exactly. The point about Ginkgo is that the only other thing that has a leaf at all like it is the maiden hair fern (well, the leaflet, not the leaf), and you're not going to confuse the two.

As an example of computer identification, it's not great. A computer program that didn't let you ID ginkgo would be totally useless. It's certainly not something I'd go berry picking with.

I'm glad someone finally noticed the meaning of my name!

Sean said it best: I think they're more sensitive to salts and chlorine, and it's easier to grow carnivores on distilled water than to accurately mimic bog water.

Plants are carnivores to obtain things like nitrogen and phosphorus. It makes sense that they'd grow in places where they've got very few nutrients and a lot of sunlight to make traps with. Mostly, they're found in bogs and in places with really ancient soils, such as tepuis and certain parts of Australia and south Africa.

I've been to four different educational institutions in my life; an undiagnosed learning disability meant I bounced around a bit. A local state university (English degree - expelled), a private university (Technical Communications - almost a Bachelor's, university slashed support for the program and cancelled evening classes, then played games with my transfer credits when I was a few classes from graduation) and two technical colleges (Assoc. in Comp Sci, and half a Graphic Design Degree)

By FAR, the quality of education I got at the technical colleges was better than the education I got from the universities. In addition, the staff and administration were much friendlier and easier to deal with. The only difference in job skill training (your core programming classes for example) was that the classes at the tech schools were taught by people who'd had to earn a living in the real world doing that kind of work.

You weren't required to take more core classes at the universities either. A person with an IT degree from a technical college has taken just as many programming classes as the person with a BS in Comp Sci unless the BSer filled out all his electives with programming.

Another example, the local tech college has an animation program essentially put together by Raven and Human Head. When they had a job vacancy open up they put out a call for portfolios from applicants. The portfolios from UW Madison Art School were terrible. Here was a group of people who sat around talking about how art made them FEEL, rather than putting their noses down and learning how to DRAW. None of them would have been allowed to graduate from the "inferior" technical school program.

People say a Bachelor's will make me a more rounded human being, but what the hell does that even mean? I know LOTS of people with Bachelor's degrees who never picked up a book after graduating and generally ignorant. I say if I want to be well rounded I'll eat more and exercise less.

A lot of businesses are picking up on this. They're no longer demanding Bachelor's degrees, but are instead willing to take anyone with certifications and work experience regardless of education level.

http://www.physorg.com/news/2011-10-low-cost-electronic-tablet-worth-indian.html

"The I-slate, an electronic version of the hand-held blackboard slates used by millions of Indian children, will eventually be solar-powered for use in classrooms that lack electricity. It is being developed by researchers at the Institute for Sustainable and Applied Infodynamics (ISAID), a joint program of Rice University in Houston and Nanyang Technological University (NTU) in Singapore. When mass-produced, the solar-powered I-slate is expected to cost less than $50 (64 Singapore dollars). "Our study clearly shows the I-slate is an effective learning tool for all students, regardless of their learning ability," said computer scientist and I-slate creator Krishna Palem, director of ISAID. "The first production I-slates will be pre-loaded with lessons for mathematics, science and social studies."

I suspect that machine learning's taken a giant step back because we know more about human minds.

I remember reading a book that assumed that humans were ultimately like flatworms, following the old story that you could teach a flatworm something, grind it up, feed it to another flatworm, and said worm would know what it had been fed. Anyway, it had a professor from Berkeley getting ground up and fed to students. No bias there...

Anyway, attempts at sleep teaching or similar speed teaching don't seem to have worked that well in real life, and so they've been dropped from the SF stable, along with, oh, the winged beings from the Cthulhu mythos, flying through the ether between planets, and guns that shot beams of mystery radiation (t-rays, anyone?). Not every disappointing technology disappears from SF, but there certainly is change over time.

Actually, T-rays might be rather nasty, esp a high energy very short duration T-ray pulse

True. Okay, let's change the example to psionic amplifiers. Or you can use whatever Edgar Rice Burroughs had them blowing out their nozzles on Barsoon and Venus.

Making a dedicated TV programme instead works, but is expensive: too expensive for us to do any more.

Damn. I think I've watched every science and engineering program you've got on iTunesU. I show them to my students (ages 15-17) because they are much better than programs made for that age group — the kids like them because they aren't being talked down to, and they don't have to filter out the fluff in 'edutainment'.

That explain why all the new stuff is ebooks, which are bloody hard to read on an iPod (and don't work right on my Kobo).

Anyway, attempts at sleep teaching or similar speed teaching don't seem to have worked that well in real life, and so they've been dropped from the SF stable, along with, oh, the winged beings from the Cthulhu mythos, flying through the ether between planets, and guns that shot beams of mystery radiation (t-rays, anyone?).

Well then, why haven't ftl, antigravity and so on and so forth been dropped? Way more implausible - and in terms of basic physics at that. Come to that, I still see flying cars buzzing around.

But mechanical education, like domed undersea cities, hasn't seemed to have made the cut into the 21st century.

There must be a way of using computers. small machines or print to teach reasoning without the use off teachers. Their time is too filled with things the office wants to really teach. I remember a magazine article on that from the 60's that had problems to solve. But you had to think, not remember.

Take your pick of explanations. One could be cultural: SF on TV typically has some form of FTL ship, and whether or not they have antigravity, they have gravity control (the better to keep the actors from working on wires simulating free fall), and it's just standard tropes for conventional writers.

The other reason is that FTL is a great way to move between scenes. Machine learning, on the other hand, has the connotation of programming. Within a story, it's best used for explaining where a character got his skills in as small a space as possible. Conversely, going to Starfleet Academy (or Hogwarts) gives the character a back story, which can be used in more ways.

Take your pick. I think there are elements of both at play.

Something along the lines of this, perhaps?

Mechanical educators/learning from a can as a SF trope is an indicator species, like artificial uteruses. Those are a marker of SF's enormous gender problem; mech-edu is a marker of the huge intellectual chip on SF's shoulder. It - we - both despise the literary establishment and desperately want it to take us seriously. Also, weren't a lot of golden age writers autodidacts who were bitter about not getting a university education?

I'd argue that the expansion of the universities since the 1960s partly explains why it died out - that and the rise of Geek Pride. Being an autodidactic programmer who likes talking squid in space is no longer something to hide, but rather something to dress up in sequins and red flags and march through the streets.

I'd also argue that there's a distinction between tropes that follow the wider market and ones that are driven by structural features of SF. Domed undersea cities = Paul Ehrlich population dread. Demographic transition = no more undersea cities. Postnuclear wastelands = Cold War. Biological horror = Terrorists/Anthrax. But things like artificial wombs, marching morons, zombies, roboteachers are system-inherent. And the robot educator is interesting as it is a system trope that died out.

I think that the reason strange learning has faded away is that some of it's been tried and failed, and it's not as interesting anymore because of that. Plus, with Wikipedia / Google / iPad wireless internet just about anywhere, who needs a billion facts stuffed into your head?

FTL and so forth are still "far enough out" to be not disproven, and also still relatively sexy compared to other plot devices / fantastic elements that might be universe enabling.

Side note - I don't personally agree that more facts in head is not extremely useful. I have several thousand books in the house and have read all but a few (most of the exceptions my Wife bought us, and the rest are on the to-do pile). My life includes significant sessions of diffuse fact and concept connection efforts, some of which have produced major innovation. You do not learn that type of learning and thinking out of Wikipedia, though it's a great resource, and you do not learn it out of University, though I learned a great deal at Berkeley.

I might have misled by saying that we (OU) don't do TV programmes any more. And we no longer, say, make 8 30-minute programmes to back up a 60-point module, which is what we did in the good old days. But we still do shorter videos which go out as podcasts, and we've no plans to stop doing these. Which is good, because they are excellent: we passed 40 million downloads yesterday. How cool is that?

We also still partially or (rarely) fully fund TV and radio programmes, largely but not exclusively through the BBC. So 'More or Less', for example, is one of ours. But these are made to be good educational telly, rather than teaching material at Batchelor's level, which is what the old stuff was.

Ebooks, yeah, fair point: the hopeful point is that they are all now written in a structured content formula which makes it very easy to update the format. People who know much more about this than me are all saying that HTML 5 will solve a lot of compatability problems with ebook formats. Hope so.

"Plus, with Wikipedia / Google / iPad wireless internet just about anywhere, who needs a billion facts stuffed into your head?"

Everyone. It's just that the particular facts to be memorized have changed. Did you know that N-acetyl glucosamine has been found to control the symptoms and damage of multiple sclerosis? I do not know the details, but I do memorize the research headlines - dozens if not hundreds per day.

I think that the reason strange learning has faded away is that some of it's been tried and failed, and it's not as interesting anymore because of that.

I find this one especially odd because of all the crap science fiction has come up with (I mean crap in the sense of Phil Dick's "The Crap Artist"), accelerated education has surely got to be one of the most useful gimmicks it's come up with.

Unlike, say, manned space flight :-)

It's also odd in the sense even as strange learning as faded, there's a demand (very well managed, of course), that "teachers be held accountable" and a notion that Kids Rnt Lurning. Trust me, as someone who's required to teach certain core ideas in an extremely limited time, if there was a way to beat them into my students heads faster I'd do it in a second.[1] Instead, it's just the same old same old; the way to get good at math is to . . . do lots and lots of math problems.

Plus, with Wikipedia / Google / iPad wireless internet just about anywhere, who needs a billion facts stuffed into your head?

Heh. All we had growing up was the World Book Encyclopedia. And we was thankful for it.

[1]I suddenly have this picture of my students shuffling in and me making sure all the connections are set up properly in what look to be suspiciously like 50's-style hair dryers. Then silence and bliss for forty minutes while they're zonked out, giving me time to read Sports Illustrated and a quick smoke break.

"If you can work on it without help, then it isn't complicated enough".

You weren't required to take more core classes at the universities either. A person with an IT degree from a technical college has taken just as many programming classes as the person with a BS in Comp Sci unless the BSer filled out all his electives with programming.

The difference is that software engineering is about more than programming. Being able to write a program is a good entry-level skill; but going beyond that first job requires a lot of other stuff. The "other stuff" is a lot easier to pick up if you have an understanding of the theory behind it.

For instance - learning all about compiler, operating system, or communications theory, may not have been much help for the first two years of my career. They were very helpful later on, when having the theoretical grounding meant that I could turn out a better solution to the problems I was hitting.

Don't get me wrong - the practical engineering experience that I've gained over the last two decades has been vital; but the "other stuff" makes you more versatile. I now work for the same firm as my old lecturer in Communications Theory; he's be the first to admit that his coding style is far from production quality, but then he's come up with some genuinely useful demonstrators that got turned into products.

Doctor Doom: yellow card. (Cause: gratuitous political trolling and unsupported repetition of crankish wingnut talking points.)

Joan, I'm spectacularly rubbish at keeping plants alive, but there's a tank on my kitchen window-sill full of nature's insecticide (in the shape of a flourishing colony of sundews and a couple of raggedy pitcher plants -- the Venus Flytrap finally died after I was away on a trip when it flowered, and they're unfashionable enough in the UK that the local garden centres no longer sell them).

Keeping them alive is mostly a matter of keeping 1-2 inches of standing water in the bottom of the aquarium, topped up from the ion-exchange filter jug we use for drinking water. They're really not that hard! Just remember they're bog plants that live off rainwater that's been filtered through sphagnum moss and they'll do fine.

A person with an IT degree from a technical college has taken just as many programming classes as the person with a BS in Comp Sci unless the BSer filled out all his electives with programming.

That's not what a CS degree is about.

(And it's why I refer to my CS conversion degree as a "Cargo cult" CS degree -- you can distil a chunk of CS down into 12 months and drop it on someone's head in a death march, but you can't make up for 4 years of higher mathematics, which is what real CS degrees give their graduates -- the theory underpinning the field, not the specifics of J. Random Fashionable Programming Language which will be obsolete in 5 years.)

Mechanical educators/learning from a can as a SF trope is an indicator species, like artificial uteruses. ... Also, weren't a lot of golden age writers autodidacts who were bitter about not getting a university education?

I think you nailed it.

Golden age authors with chips on shoulder about being an autodidact -> SFnal trope that makes those pesky professors obsolete.

Modern authors are somewhat less prone to this cognitive disorder, if only because access to higher education is vastly more widespread.

there's a tank on my kitchen window-sill full of nature's insecticide

Well, that's a good way to put it. I mostly keep cacti on my windowsill, raised from seed ten years ago from Tuscon. I haul the cacti down to the classroom when we read Dune. I inevitably get some of the barbs stuck in my arm.

Computer Science degrees are also useful for throwing random pieces of an enormous field at you to see what sticks. See: 3rd Year of my degree, where we studied (among other things) declarative programming and knowledge bases leading into AI, telecommunications, some compiler design, and microprocessor and memory caching architecture. (I'm still amazed Intel don't make more money from licensing their shared cache scheme to network filesystem vendors. It's so incredibly useful.) 4th Year varied more, but they were optional. All of us took the above. ",)

I wanted to suggest two excellent, if sobering, articles as possible starting points for further discussion here:

Neal Stephenson on Innovation Starvation: http://www.worldpolicy.org/journal/fall2011/innovation-starvation

Peter Thiel on The End of the Future: http://www.nationalreview.com/articles/print/278758

The chip on the shoulder bunch seem to have migrated to unmoderated physics forums where they constant rant on about Einstein

we still do shorter videos which go out as podcasts, and we've no plans to stop doing these. Which is good, because they are excellent: we passed 40 million downloads yesterday.

I'm some of those downloads. I use them teaching high school. I think I've downloaded every one of them, in the highest definition available.

Ebooks, yeah, fair point: the hopeful point is that they are all now written in a structured content formula which makes it very easy to update the format. People who know much more about this than me are all saying that HTML 5 will solve a lot of compatability problems with ebook formats. Hope so.

I bought a Kobo Touch because the original Kobo couldn't handle the links in OU ebooks. The Touch handles the links in terms of getting into them, but there's no back button so getting out can be tricky. And no colour pictures or video, either. An iPad is getting tempting…

"One memory trick"... not really.

Ah, that was a bit ambiguous of me — the "one" in that sentence referred not to the number of tricks, but to the number of skills — namely, rote memorisation. There may well be more than one trick to it, indeed the children are expected to become experts at it (10 years, 10k hours...) but it's still just one skill.

One skill that's at best only moderately useful in the world.

". You can learn much the same stuff anywhere. You go there to get plugged into the social network of graduates from that school."

I've been to a third-tier university, a second-tier university, and a world-class university. I can tell you that you don't learn much of the same stuff.

However, I do agree with the social network advantage of an elite school.

"Moving moving human beings around is likely get much harder very soon, for energy, political, and epidemiological reasons.

In the post-oil world, moving human beings is going to get more expensive. To get to an elite school with your specialty you may have to travel halfway around the world. "

The trend is towards more human movement, with political barriers falling. Second, moving a human being to college is cheaper than dirt, because one or two round trips suffice for an entire year. I drive 1,000 miles/month, and I don't have a long commute. Large passenger aircraft are at least twice as efficient as a single-passenger car (using passenger-miles/gallon).

For a student to go to Global Elite U from the other side of the world for a year costs about as much energy as I consume in two years of driving - and that's assuming a mid-year trip back home.

"2. Moving moving human beings around is likely get much harder very soon, for energy, political, and epidemiological reasons."

What I find interesting is that people seem to think that this will cut long-distance movement, as opposed to the longer-distance movement of the long commute (in the USA). 30 miles each way each working day adds up to quite a bit, and it's done very inefficiently. Combined with long-distance shopping and errand running, and it's hard to see that being affordable with gas over $5 US / gallon.

"The problem with modern universities is that they have allowed vast numbers of unqualified students into their ranks for the sake of various leftist political ideals, and consequently have greatly dumbed-down and politicized the academic environment. Spend some time at a secular university in the United States today; what you will find has as much resemblance to a Marxist indoctrination camp as a place of higher learning. I will therefore be quite happy to see these increasingly useless and Soviet-style bureaucracies receive the gutting they so richly deserve."

Wow. It's just amazing to watch people open their mouths and kick out whole paragraphs of utterly false boiler plate.

Check out comment #107 on the same topic.

As for transportation costs, I think the big thing that will suffer as transportation costs rise are the current international collaborations.

At Research 1 universities (the big, well-known ones), the professors do a lot of traveling, whether it's to international conferences, Washington DC for funding discussions, facilities that may be half-way across the world, research in remote corners, etc.

Obviously I've read Darwin and Malinowski--researchers won't stop traveling, even if they're on horseback or under sail. However, the number of such trips will drop enormously. This will restructure how big science is done. Since the big universities often train the professors for the smaller universities, as big programs dry up, we may see smaller schools having trouble finding qualified professors. Given the current glut of post docs, I suspect this will take decades.

The other problem is that much of the science in the developing world is actually based in first world institutions. This is a long-tail legacy of colonialism, but as such institutions contract their reach, research projects that benefit a number of marginal cultures and species will contract with it.

Moribund topic, but here's a "fun" question to consider: What if the student loan market bubble pops? Like soon? See http://boingboing.net/2011/10/07/122076.html for links.

My $0.02: it depends if widespread broadband uptake means the internet becomes a major distance learning delivery channel.

In some fields, this may be possible -- ones where specialist equipment isn't needed, just lots of study. Mathematics, philosophy, literature: all you need is a pencil, and lots of sweat. Engineering or chemistry or experimental physics or medicine? You can't teach those without serious lab/workshop equipment, or a teaching hospital. Other fields may fall in-between: what about MBAs or law, for example? What are the requirements above and beyond textbooks, lectures, and tutorials? (MBAs need face to face time because one of the things a good MBA provides the graduate with is a social network ...)

My guess is that distance learning isn't applicable to all fields, but for those where it is, many students -- especially those who just want a sheepskin to prove they studied one of these subjects -- would do a lot better to stay at home with parents and take the equivalent of Open University courses. Which aren't free, but are cheaper than attending an in-person university and having the incremental costs of living away from home.

There is probably room for some of the high-end universities to establish distance learning courses -- think, for instance, of a degree level course of study in computer science from MIT delivered remotely -- at a lower cost than in-person attendance. And once this starts happening, a lot of low-end local colleges will go to the wall (because who wants an in-person degree in Eng. Lit. from the University of Dewsbury when they can get a distance learning degree from Cambridge for half the price?)

Employers would have to start paying attention to non-academic credentials when recruiting. This is going to make life a lot harder for HR folks -- they're going to have to stop screening applicants out just because they don't have the right sheepskin or tick the right boxes -- but it would reduce the pressure on young folks to follow the credentialist paper-chase.

And we could really do with schools providing vocational skills for older but less intellectually inclined kids instead of attempting to provide an academic track on everyone.

But the invisible elephant in the living room is this: how are we going to cope in a near-future when almost all manufacturing and agricultural jobs have been automated out of existence? Much less one where robotics has reached the point where many service jobs have been automated out of existence?

Someone wrote that memorization is not what happens in post-secondary education. Well, in at least one field, Medicine, memorization is a major component of the education.

And it is a good thing too. Memorization is not everything in an educated mind, but it's not nothing either.