Five years ago I more or less finished writing "Halting State", although it wasn't published until mid-2007. Around that time, MMOs were getting an increasing amount of interest, and a startup forum/social site called GuildCafe commissioned me to write an article about the next 25 years.
So I'm reprinting it below. And my question for you is, what' did I get wrong in 2007?
Life's a game, and then you die
A shorter history of the next 25 years
I've been asked by our hosts to take a stab at identifying how online games will affect our culture over the next couple of decades. That's an interesting target because it covers a bunch of time scales. So I'm going to look at where we stand today, and where we might go at various stages along that 25 year time-line.
That's a tall order; technology doesn't stand still, and it's no good trying to guess where the gaming field is going without knowing where the tech base is taking us. So we need to look at where we are and where we've come from in order to plot a course ahead. [See boxout, "The dirty tech truth" (at end of article).]
One year out
Some time in the next year or so, I expect to wake up one morning and see a newspaper headline in my RSS reader: "TERRORIST TRAINING CELL RAIDED IN SECOND LIFE".
This doesn't mean that Osama bin Laden is a gamer, or indeed that there *are* any terrorists in SL. (Au contraire, real terrorists are more interested in blowing shit up than playing games.) Worst case: some whack jobs will figure out that SL—or it could be WoW or EO—is a cheap tool for multi-user chat that isn't currently being monitored by the feds. (Expect this window of opportunity to close about ten seconds after this article is published.) What such an article will *really* signify is that the mainstream press have finally discovered MMOs.
The media goes through a series of stages with any new communications tool or medium as they get to grips with it. And we're nearly halfway through the process.
It starts with breathless memos from the cutting-edge technology underground, reported in the media niche that subsists on sniffing out new developments. Think WIRED. "A bunch of egg-heads have invented a whizzmajig ... they're predicting global revolution!" The journalists arrive like seagulls, flying over at a great height and squawking loudly as they deliver a mess of misconceptions before flying out again.
A while later, when the whizzmajig catches on and gets a loyal user community, the first columnist notices it: "hey! You strange egg-heads in your whizzamajig! Can you hear me? Do you speak ... English?"
Then, some time later, the dam breaks.
The first symptom is that Reuters pay Warren Ellis or some other cutting-edge cyber-celebrity to move into SL. (And, whaddaya know, if they did the job right, they picked someone who actually knows what they're talking about.) Warren drinks their retainer or injects it into his eyeball or something, then dashes off some febrile prose which gets syndicated. Heads turn at AP and UPI: "why don't *we* have someone covering this Whizzumajig? We're falling behind! Hire Hunter Thompson!"
At the same time, some random gamers in places like the Swedish Foreign Ministry or the French Nazi Party decide they can get some free publicity by staking out some territory and figuratively mooning the straights. Exploding pigs, flying lutefisk, and other whackiness ensues.
And then the tidal wave of mass media awareness arrives, complete with the usual foaming mess of sewage, uprooted trees, and general crap turned out by the tabloid press and cheap news channels as they try to spew one lurid scenario after another through the playground. "It encourages pedophiles! Or terrorists! Kids get into Whizzumajig and fail their college exams! Users get hair in their palms and go blind! Ban Whizzumajigs now, before it steals our precious bodily fluid!"
This is followed by the most desperately attention-hungry members of the political class picking up the stupidest articles written by the most misguided members of the fourth estate, and proposing legislation so jaw-droppingly idiotic that their sane colleagues usually strangle it in the cradle. (See also: the internet, blogs and social network software, YouTube, MP3s, and probably papyrus back in Ancient Egypt. So has it ever been ...)
Today, about 0.1% of the planetary population have logged onto WoW at some time or another. Warren Ellis has indeed moved in on Second Life. We are therefore, within the next year or so, going to end up paddling for dear life to keep our heads above the tidal wave of public attention.
Five years out
Five years is an interesting time.
It takes roughly five years to develop a game. And by 2012, the 2G GSM phone networks that have been the backbone of global communication since the mid 90s will be beginning to shutdown. Chronically short on bandwidth, 2.5G networks couldn't push much past 100kbps, too slow to be useful for massively multiplayer gaming. Their 3G and 4G replacements, competing with WiMAX, bring orders of magnitude more data to mobile devices. Meanwhile, GPS location services are universal—for the first time, your mobile phone or laptop knows where you are and how to connect you to the internet.
Intermediate machines like UMPC's put power equivalent to today's gaming machines in a pocket. Desktop or wall-hanging TV/PC hybrids will be able to fill a wide screen to 1080p or higher resolution at fifty frames a second minimum. If true head-up displays start to appear in this time frame, things will get seriously interesting, but for now I'm not betting on it (yet). On the other hand, input devices are going to change for the better. Take the Wii controller as an example: put it together with a 60" wall-filling screen and voice-over-IP chat to your guild fellows and you've got a rather different gaming experience. Throw in some speech filters and your voice will even match your class or species.
What we're going to see is an explosion of new types of game. In addition to current-day MMOs, we're going to see systems that tie into social networks: MMOs that provide social fora similar to GuildCafe as part of the experience, not necessarily directly in-game but not out of game either.
And then there are the kissing-cousins. Modern MMOs are largely descended from table top role-playing games such as AD&D: but they're not the only spin-offs of those original paper-and-dice passtimes. Live Action Role-Playing (LARP) involves getting out and about and playing your role in the field; at risk of over-generalizing, LARPing bears an equivalent relationship to RPGs that paintball games bear to first-person-shooters. However, while paintball has tended to involve technology all along, LARP has mostly not been infected with spurious electronics. All of that is about to change, though, thanks to the tools coming down the pipeline over the next few years.
Geocaching is a seemingly unrelated game; in it, players with GPS units compete to track down cached treasure troves containing cryptic clues, following the trail wherever it leads. Geocaching and live-action roleplaying haven't crossed paths with each other much so far, or with MMOs, but that's going to change. With UMPCs and GPS, your MMO can track players in the real world. This is going to evolve slowly, growing out of LARP routes rather than from the MMO side—but once wearable displays show up, live-action MMOs will soon follow.
An interesting bolt-on that is lurking just over the 5-year horizon is the cheap motion capture suits. Used for developing games by digitizing actor's movements, these suits are currently horrifyingly expensive: but they're going to get cheaper, eventually imploding into a bunch of small boxes on velcro straps that you attach to your limbs. Once you're wired for motion capture, the game doesn't just know where in the world you are—it knows whether you're parrying or picking your nose.
Given this choice of gameing paradigms, do you stay at home with your big screen and keyboard (or if you're feeling more energetic, your motion cap sensors) and game with friends on another continent, or do you go find a muddy field and game with friends who live locally? Well, why not do both? If you've got head-up displays (and there's no reason not to equip them with cheap camera chips so you can dodge real-world obstacles mapped into gamespace, as well as your view of your MMO), then there's no reason not to mix live action games with telepresence.
But it'll probably take a couple of generations of game before the mixture of long-range MMO and VR-assisted LARP really matures. Watch the skies.
As for the number of players ... I'm going to bet on an order of magnitude increase over the next five years. That's right: 1% of the human population of this planet will have tried MMOs by 2012. That's about 65 million people. A large proportion of the new players will be Chinese, Russian, or South American, but this isn't necessarily bad—there'll be genuine players showing up, not just farmers. Expect some interesting fallout (and possibly organized large-scale griefing) as real-world cultural divides come to the fore.
There are going to be non-game side-effects, too.
One of the big new things bubbling up under the tech world is 3D printers—machines costing anything from a couple of thousand dollars up to a couple of million that take in a CAD diagram and some raw materials and spit out a finished product, be it a wood carving or a piece of molded plastic or an extruded concrete house.
This isn't a particularly new technology, but it stayed eye-wateringly expensive until relatively recently. Now there are DIY plastic extrusion machines available for just over $1200, and the field's waiting to take off, much as PCs were in 1975.
As and when 3D printers catch on, a lot of hobbyists are going to be very happy—as well as architects (think: models), engineers (who need odd-shaped widgets) and DIY enthusiasts (you want that shelf bracket to *really* fit, design it yourself). But most of us will be content to buy or download open source templates for objects, possibly tweaking them to fit our specific requirements before we print them. And what better environment for seeing if that shelf bracket really *is* suitable for the living room than a virtual reality environment? MMOs are the first commercially viable VR systems, and the less-game-oriented ones are going to be moved in on by bespoke retailers—as opposed to off-the-shelf vendors like Amazon or WalMart—once they, and 3D printers, become widespread. Instead of going to the furniture store and puzzling over whether that sofa is going to fit, you can preview it in your Second Life living room then order it online for real-world delivery. And for smaller knick-knacks in SL (or whatever its successor is in 2012), you can just hit "print" and wait for it to rise up out of your 3D printer.
This is going to have huge long-term effects, incidentally—although it'll take a lot more than 5 years for them to become pervasive (think more like a generation or two). We're used to most of our big purchase items being mass produced by factories, in a limited range of shapes and sizes. With the level of customization that VR previewing offers, never mind the advent of 3D printers, a lot of items that today we buy from mass production runs are going to be so heavily customized that they might as well be bespoke originals. The standardized "look" that so many furnishings and domestic items have, encouraged by huge chains like WalMart or IKEA, is gradually going to become a rarity, replaced by a much more diverse range of products.
10-15 years out
Do you remember what the internet was like in 1992? Or 1987? Those are the years that bracket the 10-15 year range, looking backwards instead of forward. At the earliest reach, in 1987, the first MMO games just about existed—the first commercial, graphical MUDs showed up in the mid 80s. Trying to predict what a 2017-2023 game will look like, from here, is a losing proposition, like trying to predict Burning Crusade from the perspective of a user of the Exeter University MIST system.
So let's look at the development picture instead.
One of the problems with MMOs is that we want lots of lovely eyeball candy, cute tiles and interesting loot and fun scenery to gaze upon. This is getting to be a major, even dominant, aspect of the development process. As of 2006, typical games cost millions to develop—and may employ anything from dozens to hundreds of artists. If creation of new content isn't at least partially automated by 2023, the entire gaming biz is going to be in big trouble, with production and maintenance costs skyrocketing.
Procedural content—for example, automatic dungeon design—is nothing new; Moria and Nethack and similar games were doing it in the eighties. But they didn't have to handle graphic design. We should, by 2023, have some serious parallel processing hardware, with workstations containing up to 1024 cores per processor (and processors hopefully running at better than 10GHz by then). One possibility is to use genetic algorithms to evolve new character templates and monsters; Will Wright is exploring this idea (in a different context) in Spore and it's not hard to see it used to run variations on a theme in a more traditional MMO.
Another likely war forward is user-created content. Looking at the user contributed content for Neverwinter Nights, it's hard to separate the game's success from it's players' enthusiasm for DMing. One thing that's certain is that any games company that wants to make a profit is going to invest heavily in DM and content creation tools. The first game company to figure out how to recruit and reward users for contributing content and admin skills to their realm is going to win big.
Now here's another point: how big can we make an MMO shard?
Right now, it's hard to put more than a couple of thousand players on a single shard and maintain a coherent multi-user reality. There are bandwidth problems and server problems. But we're seeing three tech developments in 2007 which might help with this situation: peer-to-peer networking, trusted computing with DRM (by which I mean, someone else holds the keys to some of the services on your computer), and distributed databases.
We take it for granted that players *will* try to hack any system they have access to, for whatever purpose. (Especially if real world money is passing hands, exchangable for in-game goods and services.) At first glance, the idea of doing away with the current model of central servers and players running a client program, and replacing it with a peer-to-peer model, looks like madness. But if you can come up with an authentication and encryption scheme that uses the DRM and trusted computing hardware built into their computers to stop them from tampering with a secure server partition, and if that server is then talking to other player's game clients, sort of like SETI@home for MMOs, then where can we go from there?
Potentially we can get away from the scaling problem that has bedevilled MMOs—Ultima Online used to have to add a server for every hundred new players, eating into their profit margin immensely—by getting the players to subsidize the servers, and to contribute the content, and even to do some of the admin work, a game company could focus on improving its bottom line through sophisticated manipulation of the exchange rate between in-game goods and real world money. Of course, derivatives trading is risky—someone loses money if someone else makes a killing. Expect this one to appeal at first to MBAs and then to computer criminals everywhere. In fact, by 2023 some hideous financial instrument lurching out of a computer game and crashing an economy the size of Hungary will probably replace the Ponzi or pyramid scheme as the acme of web-of-trust scams.
(Random prediction: sooner or later, some enterprising crook will try running one of these scams on a bunch of hardcore LARPers. And they'll run into a world of hurt when they realize that *these* marks expect to get up close and physical, unlike the more virtual kind.)
A less crazy business scheme might involve giving away a p2p game environment, on condition that the players sign over a chunk of computer time to the games company—who can then resell it. Can't imagine who'd want 1% of the CPU cycles of ten million servers? Consider the usefulness of such a service to biotech startups trying to do protein folding calculations, or to organizations that wants to do parallel searching on terabytes of data—Google, for example (or the NSA).
By 2023, I wouldn't be too surprised if 750 million people have tried MMOs—that's 10% of the world population. But by then, growth will be slowing. Only 20% of the population have access to PCs, and even by 2023 I doubt that much more than 40% will have mobile phones. So the market is going to be pretty much saturated by then, and anyone hoping to make a killing is going to have to come up with something new to sell.
On the other hand, there'll be a lot of business opportunities in an ecology where MMOs are ubiquitous and games companies make money by selling add-ons. The flip side of "how big can we go?" is "how small?" Odds are high that personal islands in the successors to Second Life—inaccessible without an invitation, secured by biometric dongles—will be something we all have. And we keep stuff there that we don't need all the time but might want to drag kicking and screaming out of the virtual world and into reality. Our business avatars, for example: motion capture and VR finally answer the marketing problem that's bedeviled video phones, "do you want your boss to see what you look like and where you are when you answer the phone at 3am?" And our business clothing: with vastly more sophisticated design tools and a plethora of bespoke retail outlets, most of your wardrobe could well be virtual, only turned into actual fabric (by robots, with courier delivery) if you decide you actually need it. Want to go on vacation for a month? Pack a change of underwear and head off—you can bounce your wardrobe contents at a local fabrication store and pick up extra stuff as and when you need it.
There'll be more subtle legal and political changes, too. Right now, DRM—digital rights management—is a dirty word to most consumers, and Trusted Computing is seen as a euphemism for greedy software and hardware monopolists wanting to lock down our computers. But once large chunks of our actual public identity exists in virtual realms, we're going to badly want these technologies (with biometric authentication on top) in order to keep control of our virtual selves. Our understanding of "intellectual property" is a movable feast, and once we get used to stuff inside MMO environments not only looking like "real stuff" but being exportable into the real world, it's going to change again.
15-25 years out
MMOs are today (2007) the first really commercialy successful form of virtual reality. And the internet is something that exists inside our computers and mostly doesn't erupt out of them at random intervals.
By 2027-32, this position could well have reversed. For starters, the display tech we'll be using by then may well be built into our eyes, or at least into unobtrusive spectacle frames. (I'd bet on the latter—but then, I can't stand contact lenses either.) And for another, we'll have run down the off ramp from Moore's Law. Game over: we've hit the nanoscale, our machines are the size of molecules, and the limits to computing will be more about heat dissipation and signal paths than whether we can build a higher resolution fab line for two-dimensional chips.
Our computing resources will know where we are and how to find us. It'll be able to see us and read our expressions and sense whether we're tightening our muscles and precisely how we've positioned our feet. And anything on the net that we've authorized to look at us will be able to see us, too. Or see our avatars. It'll be a lot like the pre-modern concept of a spirit world, except we'll have access to it whenever we need to and it won't be asking for burnt offerings. The really interesting question is whether things will converge on a single overarching metaverse with games or business meetings happening in different places, or whether they'll fracture and we'll see even more divergent environments cropping up.
This tech isn't just going to stay the domain of gamers. Probably in as little as five years time it's going to be in use by other folks. Think of your local police department as playing an elaborate game in Copspace—a VR environment overlaid on a geographical map of their territory, marked up with case files and notes, with a live video evidence feed coming off the cameras on their badges. Or think of your own lifelog. Storage is so cheap that you can record everything that happens to you, using voice annotations to mark points in time so that your phone can later recall them for you.
Whether you live and play in augmented reality or virtual reality is a choice you'll make every day. Probably the two environments will overlap so that the next generation—the folks born in 2012—won't necessarily be aware of it. Do you remember growing up before computers? Before CDs? Before GPS? Before the internet? They're not going to remember growing up before MMOs, or VR, or AR. The politicians grandstanding today about the evils of computer games and the urgent need to ban Whizzumajig will look as quaint to their eyes as a Prohibition-era preacher ranting about the evils of the demon Rum.
There'll be more subtle side-effects, too. GPS is squeezing in everywhere, not just into games. We've already seen the first swarming open source mapping runs, when volunteers get together with GPS and mapping software to create detailed open source maps of an area. By the 15-25 year time frame, we're going to be forgetting what it's like to get lost by accident. Your phone (and your avatars) will always know where you are, and more importantly, how to guide you from where you are to anywhere you might want to go. Paper maps will follow log tables and slide rules into the tar sands of definitively obsolete technology, and a whole slew of movie plots will become implausible (in the same way that the cellphone blew most teen slasher film plots out of the water). Of course, if you can corrupt someone's map of the environment they move in, new types of crime become possible: imagine muggers who guide their victims astray into an off-camera blind spot, rather than stalking them through surveilled streets.
And of course, all of this is wrong.
Life's a game, and then you die
Gaming is meant to be fun. What I've been exploring isn't fun; it's the consequences of new technologies for game-play erupting into the real world.
Our idea of recreation and amusement changes over time. Today's drudgery is tomorrow's entertainment, and vice versa: table-top wargaming was a hardcore exercise for the Prussian general staff in the 19th century, while putting in several hours of Polo on horseback was mandatory for British cavalry soldiers in India—it was excellent close-order cavalry training. A lot of today's occupations would look like a game to an earlier generation, and vice versa.
One constant over time seems to be that each generation invents its own fun. MMOs are the current generation's big thing; back in the 1970s it was paper-and-pencil RPGs. By 2023-2027, it's reasonable to expect something else to come along, something that pushes the same buttons and is even more absorbing. But what will it look like?
That I can't tell you. All I know is, it'll be fun.
BOXOUT: The dirty tech truth
I'm writing this in early 2007. I'm sitting in front of a laptop with a 2GHz Core Duo processor, 2Gb of RAM, and about 0.7 terabytes of storage plugged into it. My net access comes courtest of either a 10 mbps cable modem or an early 3G mobile phone that can reach the dizzy heights of 384kbps.
This is pretty typical for early 2007. But to put it in perspective, back in 1983 the entire planet bought just 93 terabytes of storage and most cutting-edge PCs were running on 4.77MHz 8086s with maybe 128Kb of RAM and an 0.3kbps modem. Those 4.77MHz CPUs took more than one clock cycle per instruction. So they were probably running at about 1MIPS. And I'm not going to get into their graphics cards and 3D accelerators and physics coprocessors because, um, they didn't have any. Let's just say, you wouldn't want to run "World of Warcraft" on an original IBM PC 5150.
So here's my expectation for tech: if this goes on, in just under 25 years someone with a reasonably current computer will have the equivalent of 1% of today's planetary data storage on tap. Their processor will be about two thousand times as powerful as today's best, and their mobile phone (or equivalent) will have more bandwidth than today's best cable modem providers can serve. And at least one of these predictions will turn out to be laughably wrong because some currently unforseen development will have upset the trundling apple cart of progress that Moore's Law has trained us to expect.
My personal favourites for those disruptive technologies? Pick any of these three:
Firstly, a year or so ago Intel finally realized that if they kept trying to burn their way through performance bottlenecks, pretty soon their processors would need asbestos underpants: so they switched from emphasizing clock speed in gigahertz to looking at how they could deliver more MIPS, more raw computing operations, per watt of power. We've already got pretty powerful embedded ARM chips in our mobile phones—with 80Mb of RAM and a 300MHz processor, and a 2Gb memory card, my phone has more raw power than my desktop did back in 1997—so I think we can expect typical phones in 2017 to be as powerful as a high-end gaming PC in 2007. As for gaming PCs, they're going to fission—you'll have laptops, and you'll have TV sets/home entertainment centres with 60 inch screens that just happen to also be PCs with capacities to rival today's data centres.
Secondly, we're getting GPS (and the European Galileo) service everywhere, and it's getting more precise and more ubiquitous. (More on what this means when I get over my nerd-tech exposition.)
Thirdly, our communications bandwidth is going through the roof. I've gone from 56kbps to 10mbps in ten years—three orders of magnitude increase. It's only going to go higher. Early tests of hardware implementing the draft 4G telephony standards managed to get that much bandwidth over the air, and as for cables, 100 gigabit ethernet is already off-the-shelf if you've got enough money. There are hard limits—you can only cram so much data into the radio spectrum—but it's going to take us a long time to reach them, and then we'll be looking at terabits per second. At which point, somewhere before 2017, it should be possible to download enough data over a cable in *one minute* to fill every byte of hard disk space sold in 1983. And the wireless tech will catch up maybe five years later.
But remember, these are just guesses, folks. I may be an SF writer but I'm not a prophet. The disruptive tech could be quantum computers, or AI, or magic wands. I'll just be very, very surprised if nothing shows up at all.