The Incredible Secret Future Of Videogames

This is a heavily revised version of an article first published in PC Gamer UK, last year. It’s based on a presentation given by Ray Kurzweil at GDC 2008, and subsequent conversations I had with the author Charles Stross and the game designer Eskil Steenberg.

This article began back in 2008, when I was sat in the audience for GDC’s keynote speech by futurist Ray Kurweil. “Games are the harbinger of everything,” Kurzweil was saying, as he delivered his take on the future of everything (with slight emphasis on games) to the assembled design-masses. The controversial technologist, who regularly talks about his hopes for technological immortality and transhuman ascension via artificial intelligence, was arguing that games were where the future manifested itself. “Ultimately,” he said, “they’re going to be competitive with real reality.” These were claims I’d heard a bunch of times before. Often, in the pub, during conversations with Kieron, as he gesticulated dangerously behind a glass of wine. Other times from developers, writers, and gamers. I realised, watching Kurzweil, that it was something I was going to have to write more about: just what kind of future do games promise?

Futurologists and science fiction writers have long been making claims about the role of technology in our lives, either speculatively in futurological texts, or simply as entertainment, in science fiction films, books, and comics. However, they’ve only recently begun to take videogames as they actually exist in the real world seriously as a major part of our future. Look back to 1960s visions of the future, and games aren’t even a twinkle in the futurological perception And yet in terms of the reality of future-becoming-present that we actually experience they are a major factor. They’re one of the most visible symbols of both technological progress, how that progress is influencing our culture. Games are a hybrid thing: a fusion of art and science, and as both disciplines push onward, so they push games ahead of them. They’re the construction yard in which many of the people who are working on the future – designers, occasionally writers and artists – are to be found.

One such writer, a man who is regularly questioned about the future, is Charles Stross. He is the author of a number of books set in the near future. He’s not here to predict the future, he stresses, but to write fictions that reflect on the present. “The book which, I think, is the reason I get invited to talk at computer games conferences and similar, is Halting State,” says Stross. “It’s a book about skullduggery in the computer games industry, and in particular about the future of MMOs.” Halting State doesn’t sound all that fanciful when you consider the kinds of scams and gold-farming exercises that already take place within the real world. Stross’ book sees a crime inside a futuristic virtual reality MMO linked to real world espionage and crime. These days that could almost be a PC Gamer news headline, particular when you consider how real-to-game currency systems have already been rumoured to be used for money-laundering for organised crime.

For Stross, the catalyst of his imagination is the current trajectory of technology. Right now he’s excited about the implications of rapidly ballooning bandwidth. “If you go back just a few years WiFi hotspots were rare, and data on a mobile phone cost extortionate amounts of money,” says Stross. “Nowadays it’s got much cheaper, and this trend is accelerating. We’re seeing the roll out of 4G, which we can expect very high data rates, which means we can expect to see mobile devices which have the equivalent bandwidth of full bore WiFi wherever we go. And it’s going to go a lot further.” For gaming, you can already achieve the kind of latency required to play Left 4 Dead over mobile broadband. That is only going to become easier, faster, and cheaper.

But Stross sees another trend too, a blurring of real and game worlds. He argues that the sea-change will come with “ubiquitous location services.” Those systems that can pinpoint our mobile phone down to a single street. “We haven’t quite gotten our heads around the idea of having devices on our person that always know where we are,” says Stross. “This is less obviously gaming related, until you start thinking about augmented reality, or live action role-play. You can play games in the real world without having arranged to meet. If you’re in the same area as another player of one of these games, for example, your phone could steer you towards each other, so that you could interact.”

Game developers are already working with this idea. London-based Mudlark are making a mobile phone game in which players use GPS to try and walk rings around each other in the real world. Which means a player stuck at his desk might really get in trouble…

Of course this kind of stuff comes as little more than augmentation to existing strata of gaming tech, as Stross observes: “It’s not going to render the tiers of gaming we already see rendered obsolete. What it will do is allow for a greater number of games to be possible: live action roleplay, spatial location games, MMO’s accessed from your mobile. These won’t replace high end PCs or consoles, but they may tie into them, or act as alternatives when you’re on the move.”

The future of games, according to Stross, is mobile, connected, multitudinous.

And if Ray Kurzweil’s speech is to be believed then we potentially face not simply expanded connectivity, but a billionfold increase in computer power in the next ten years. This has astonishing implications for the way games are going to be rendered, and delivered. Combine this leap in power with Stross’ observations about bandwidth and we suddenly get a clear glimpse of what the future could hold. Last year’s big storm of ideas: cloud-gaming. Games rendered remotely and then streamed to your PC, your laptop, or even your phone. It’s already happening, albeit with plenty of problems along the way.

The idea, as you are probably aware, is quite simple: processing power could be centralised in tohuge banks of professionally maintained servers – an easier option than for us to be continually fiddling with processors, graphics cards, and drivers (if less fun, as I would miss my tinkering). If services like the game-streaming OnLive are to be believed – and their claims are matched by the claims of futurologists – then the system you browse the net on will be all you need. As long as the bandwidth is there and your system can decode some high-res video, it’ll be all you need to play increasingly sophisticated, ultra-high-fidelity games of the near future. The game will be “played” on a rented machine at a secure location, and even cheap laptops will be transformed into awesome gaming machines.

But all this does little to actually change the future of games. It’s little more than an alternative distribution model, and that’s neither convincing, nor particularly exciting. No, there’s something else more significant going on here. Like every other rapid progression of technology, increases in processing power won’t be developing in isolation: there’s also the implications of things like nano-tech, bioengineering, and materials science. Because of this Kurzweil’s predictions, even more so than science fiction writers like Stross, are outlandish ones.

Kurzweil argues that the progress of the computer is a progress inwards. It’s going to become increasingly part of our our biology, and that means it’s going to become part of what it is to be human. Continuing the “smooth curve” of our cybernetic integration with things like mobile phones (which, in practical terms, render us telepathic) or even wrist watches – which firmly position us in time – the future will deliver more and more tech into the body. Within a couple of decades nanobots will be in our very blood, says Kurzweil. These will allow us to experience games in “first-person”, where our very eyes are the screen we carry around and use. Reality, he says, could be reduced a window in the corner of our specially augmented consciousness. People chuckle about this in the GDC audience. It’s too ludicrous to be taken seriously, obviously…

By 2029, claims Kurzweil, AI will surpass the human brain. Artificial intelligences will be joining us, will have already joined us, in games worlds, and the real world. Even before that near-human intelligences will be making all aspects of our life easier, and our time in games more enjoyable. They will almost certainly be the stars, or even the co-creators, of the future generations of gaming entertainment.

“It’s the holy grail of much of game development,” says Stross, when I ask him about artificial life and the idea of wholly created people. “Character development is even harder than building environments. What you need to be able to do is not necessarily recreate human intelligence, but simply to convince players that what they’re interacting with is interesting.” If it moves and barks like a dog, people will treat it like a dog, even if it’s on a screen. The same is going to be true of people.

“People are interested in people, at least when they’re not shooting them,” says Stross. “MMOs are popular for this reason, they become more interesting the more you play them, not because the game gets more interesting, but because your relationships with other players becomes more interesting. It generates incredible loyalty.” (Could you be loyal to an artificial person? Could games create AI that were tuned to be friends or, perhaps even better, enemies you’d be motivated to fight with?) MMOs of the future must aim to be places that are populated by both human and artificial intelligences.

The crucial issue here, as Swedish coding-mastermind Eskil Steenberg points out, is that AI can be asked do a lot of things that that people cannot, or will not. “You can’t get people to play all the boring parts in a game like guard #324,” says Steenberg. “But what you can do is to have the game makers record what the guard should do. The obvious problem with this is that the recorded actions of the guard cant react to anything unusual the player does. The solution is to program an artificial intelligence, that can control the guard and that can react to what the player does. The smarter the program is, and the more things it takes in to account, the better the experience will be for the player.” Artificial intelligence, and not pure power/bandwidth muscle, is where the true germ of our gaming future will lie.

This seems obvious: elements of artificial intelligence are popping up everywhere in games, from enemies in Halo to your neighbours in The Sims. But Steenberg takes it further. If artificial intelligence can bring characters to life, why can’t it start bringing the rest of the game to life? “If we look at other things in games, like story, levels, characters, events, cinematics, conversations they are all provided on the disc,” says Steenberg. “They have the same limitations of the guard with the recorded actions, they cant react to what the players do. The two solutions would be to either have people making all the content while you are playing the game (something that is impossible to do fast enough), or just like in the case of the guard, you write a little program that does it for you.” And, to a small degree, we’re already seeing that happen with Left 4 Dead’s director. In the sequel, it will even change the layouts of the levels. It’s also happening in Steenberg’s own game, where the AI settlements make war on the player-constructed outposts.

“Procedural content is not random content,” says Steenberg. “Just like the AI actions on the AI guard aren’t random. It follows a set of carefully set up rules that governs its actions. The more things you take in to account the better the content gets. We are seeing some nudging in this direction: destructible environments, ragdolls, physics, and AI directors. All these technologies remove the need for canned content and replace it with a bit of code.” Eventually, perhaps, games will be building themselves ahead of us, responding to us with intelligence, with creativity. Whole systems that are generated by our actions. They won’t be static, nor any more predictable than the real world. (The butterfly effect becomes a game-design principle.)

“Ultimately,” says Steenberg, “we can replace more and more of our game content with code, and we will get more and more flexible games that can react to the player in entirely new ways.” For this far-sighted Swedish developer, what is exciting is not new hardware, but software solutions devised for the hardware we already have. “I would argue that technology has digressed gaming,” says Steenberg. “The higher you put the bar, especially for graphics, the more you sacrifice in gameplay. In a 2D block based game like Super Mario Bros, implementing destructibility is easy, while in a modern 3D game it is hard so most people don’t bother. Another example is AI. Halo 3, had to scale down their AI for Halo 2 because the added graphics made it impossible to have the same number of visibility tests and path finding tests. In a text adventure it was easy to let the player chose his or her own name and then have it be pasted in to all the dialog, in a “modern” game with recorded voice, you simply can’t do that.”

Steenberg argues that the future of games is one in which software will have to find solutions for the enormous problems that following the curve of increasing hardware sophistication has presented us with. “The examples of how things that used to be simple have now become hard are numerous,” he explains. “Dwarf Fortress and similar games give a hint to where games would be, if graphics and sounds didn’t stand in our way.”

So perhaps despite the technology, and the better visuals, the march towards photorealism and perfectly rendered worlds is actually opposing progress. Perhaps the coming decades of gaming lie down another path, a path of smart AI and communal, multiplayer story-telling where humans and AI mix to make their own gaming experiences. Stross had his own observations about this: “When you see DC and Marvel comics on the shelf, we don’t pick them up expecting them to have made progress towards photorealism, although given computer tech and photoshop tools, they could go for it if they wanted to. What they’re looking at is a better way of telling a story. The ability to generate realistic 3D visuals is a red herring in terms of representing progress, at least if we’re looking at a changing game experience.”

And so we come back around to Kurzweil’s speech. “Play is how we principally learn and principally create,” said Kurzweil to the game developers. In saying that he points to another reason why the coming decades of gaming might be more about how we interact and how we play, than about how hi-tech the platform is. It’s the exact reason why the billion-fold increase Kurzweil predicts might not be the key factor in our gaming future. The key factor will be our own interest in creativity. We have to want games to do new things.

Stross agrees on this point, highlighting the crudest of gaming technologies, and their sophisticated results: “The reason pen and paper RPGs don’t die is that it’s about consensual story-telling. It’s about creativity. Most people don’t want to be creative, but one in a hundred people want to get involved: they’re bored by passive consumption. Consequently, they’re willing to pay for more stuff.” This is backed up by the revelation that MMO players actually buy more single player games than other gamers. How much, indeed, do pen and paper obsessives end up investing in their hobby?

The point, it seems, is not that games are “going to be competitive with real reality,” but that games are going enrich “real reality”, and be interwoven with it. “Virtual reality” was never a good term. As American writer Steven Shaviro points out, it should always have been know as “prosthetic reality”. Games are going to extend our reality, and they’re going to do it by being smarter, and more varied. The future is gaming heading off in all kinds of different directions. There will be no single, defining trend, other than their diversity. “The global turnover of the games industry is about $40bn, which is comparable to Hollywood,” explains Stross. “And yet we don’t think of Hollywood as a single monolith of “the movie industry” as we do can tend to do with “games industry”, we don’t confuse Pixar with slasher horror movies. We wouldn’t expect directors and producers from one area to have any knowledge or even familiarity with another area. The same will become true of games.” Stross continues: “Games have really not been around that long, and it’s still possible to be genuinely revolutionary. Imagine, by comparison, being an engineer working for Airbus or Boeing. You’d be shaving percentages off fuel efficiency, and never having a chance to be genuinely creative. That’s not true of the games industry, which still has an exciting future ahead of it.”

That future might not even involve the hardware changes that the big manufacturers are trying to push on us, if Steenberg is to be believed. “Technology doesn’t really matter in my opinion, I would be very surprised if people don’t play games with mouse/keyboard, and joypads in twenty years time. Despite the ‘one console dream’ we are getting more and more platforms that are more diversified then ever. So in conclusion I say that there is no one future, there is lots of them, and to me that sounds very good.”

I tell Steenberg about Kurzweil’s vision of a super-AI mediated through blood-borne nanbots. “Do I win anything, if I am right and he is wrong in twenty years?” Sure, how about shiny badges with RIGHT and WRONG on them? “StarTrek badges?” says Steenberg. “I’m assuming we will all be working on a starship for the Galactic Federation in 20 years…”

Maybe we will. This much is certain: there’ll be games on that Starship, wherever it goes.


  1. Tei says:

    I have not read the whole article (that looks like full of substance, and It will be a delicious read), but I will make a pre-emptive attack^H^Hcomment.

    You can write games in a farm, on the countryfield. Most of the task of creating games is manual, game devs are mostly artisans. Creating games seems slow, painfull and long. It takes patience, endurance, strong will and artistic skills of different people. It looks to me a lot like building cathedrals. Moddelers and mappers and cathedrals can share ideas and solutions…

  2. Lobotomist says:

    “Dwarf Fortress and similar games give a hint to where games would be, if graphics and sounds didn’t stand in our way.”

    I Very much agree with this. (Although i am partially responsible as graphic artist by profession.)

    Graphic (and high value production) pretty much destroyed todays games.

    I wish we throw away the shackles and return to more interactivity and gameplay opposed to eye candy.

    • Sobric says:

      I actually disagree. Sort of.

      I don’t think games are “destroyed” by their flashy visuals, far from it, they are often much improved. The problem lies in developers concentrating solely on visuals while neglecting other aspects.

      Some games benefit more than others from great visuals too. Far Cry 2 was such an atmospheric shooter largely because of the visual and audio presentation. The upcoming Battlefield Bad Company 2 is a wonderful explosion (literally) of visual and audio, which makes the whole shooting-people mechanic (which, let’s face it, isn’t particularly original) a lot more refreshing and exciting.

      The trouble is, I suppose, that too many developers spend too long on creating their graphics engine that other aspects of the game suffer.

    • Stromko says:

      A good engine and excellent presentation can substantially improve the user interface, if they are integrated and thought out properly. There’s just too many cookie-cutter ‘AAA’-quality games spending too much time and energy on state of the art graphics and environments.\ I have no doubt that Dark Void would have been quite a bit better had they neglected graphics in favor of tuning up the game itself.

  3. Alexander Norris says:

    A glass of wine? What kind of a tosser drinks wine in a pub? :P

    And if Ray Kurzweil’s speech is to be believed then we potentially face not simply expanded connectivity, but a billionfold increase in computer power in the next ten years.

    It’s the same old claim, isn’t it? “In the future, we’ll have flying cars and hoverboards and jetpacks and robot butlers, oh my!” Unlike Mr Kurzweil, I’m not in the business of predicting the future – but I very much doubt we’ll see super-smart, lifelike AI any time soon. Sure, they might be able to code increasingly complex AI that can fool the player in more and more ways, but it’ll always end up betraying its nature when the player does something unexpected. Until we reach the singularity and AIs can self-improve (in which case we’re fucked), game AI will only ever be able to do what the programmer thinks it should be able to do.

    As Jim mentioned, we’ve been promised virtual reality and the holodeck for quite some time now, and we’re still miles away from it. That said, augmented reality is a very real possibility, and so are games that interact with that. The question is, will gaming in the 2030 or everyone-is-always-connected-at-all-times look anything like what we think of as gaming (i.e. modern FPSes/RTSes) or will we all have moved on to something that looks a lot like a gigantic ARG mapped over real life via HUD-displaying contact lenses? I’m not so sure there’d be a place for bog-standard video games (as we understand them now) in this fabulous new world of hyperinterconnectivity.

    • Kieron Gillen says:

      Alexander: One who can’t drink beer any more.


    • Lobotomist says:

      Actually he is quite hardcore. He drinks wine in between his beers

    • Ffitz says:

      I’m not so sure there’d be a place for bog-standard video games (as we understand them now) in this fabulous new world of hyperinterconnectivity.

      I’m damned sure there will be a place for bog-standard video games in 20whenever.

      Two reasons: a) many, many people will be too fat and unfit to be able to run around in an augmented reality, unless the necessary equipment comes built into a bellywheel, and

      b) Running around the streets of with a plastic gun peripheral for your ARG will get you shot by the local plod.

    • Heliocentric says:

      If we do get sentient ai you can forget playing games and you can focus on mining iron out of jupiters moons under robot tyranny.

    • mandrill says:

      You really need to read/watch/listen to Kurzweil’s body of work before saying “Bah its all flying cars and jetpacks again, it’ll never happen”.

      His arguments are based on the exponential growth of technological ability (not just computing power, though that is a major factor). Best explained in this talk on TED: link to

      Its not just Moore’s law (which relates to the number of transistors on a chip) but a whole range of other exponential growth curves. Bandwidth being just one.

      The singularity is coming people, and in our lifetime. Be prepared for a big surprise (if we survive that is)

    • AndrewC says:

      Computers would calculate that it would be very inefficient to create these nightmare scenarios we think of – like being forced to mine ore and that. A specially designed robot would do the job much better. We would have to hardwire such human traits as sadism and revenge into these machines for them to act so thoroughly imperfectly.

      No, what our AI overlords will probably make us do is play computer games on them all day, an activity that would feel, to them, rather like being lightly tickled, or like stroking a hamster.

    • Ballisticsfood says:


      Yeah, but at least I’ll be top of the ore-refining and breathing-in-toxic-gases leaderboards.

    • Sobric says:

      @ Andrew C

      If that’s the case, then I welcome our robotic overlords!

    • SheffieldSteel says:

      Alex, the difference between having expectations such as robot butlers / flying cars, and on the other hand, a revolution in cyberspace, is that flying cars etc demand the efficient application of existing or new technologies within the constraints of physical reality – in other words, an engineering challenge. Inside a computer, by contrast, anything is possible. Let me qualify that. Short of possessing a soul, or exhibiting free will, a computer-generated world is limited only by our imagination and the effort we’re prepared to put into building it. One could argue that a flying car is just as feasible – it being only a matter of cost – but manufacturing costs are essentially zero for programs and data, which means that a desirable product can be sold in unlimited quantities – assuming it is desirable enough.

      Okay, that’s two differences.

    • Alexander Norris says:

      @mandrill: I’ve watched that particular TED talk and read around a little about Kurzweil, so I’m not exactly dismissing everything he has to say off the cuff (though I’m by no means an expert in any of the fields needed to make accurate predictions about the future) – I just very much think that games that are “as real as real reality” are a pipe dream, much like flying cars and space colonies on Venus. They’re something which we conjured in the past, and which is no longer really relevant to the future of gaming. What is relevant is the high levels of connectivity that we’re graduating towards; hence why I think the future of gaming lies in twitter and iPhone apps rather than “cinematic” AAA titles with photorealistic graphics.

      @SheffieldSteel: There are still costs associated with lifelike AI that make developing it as prohibitive as developing flying cars or robot butlers. Assuming that Moore’s law holds true and that we’ll eventually have the hardware to emulate real life on a computer, my argument is that it isn’t something that’s going to happen any time soon. AI still has to be programmed, which means it’s still subject to the same limitations as its programmer; and until we become a post-scarcity society where everyone has nothing but leisure time, these limitations include the need to make a profit on the sale of their game.

      Until we can develop a programming language that makes it possible to make vastly complicated AI by just talking at a computer, and programmers can be paid to sit around all day thinking about all the things that a player might possibly try to do and then just tell the computer what to do as a response, AI complexity is still going to be limited by time and effort, because your game ultimately has to sell enough for everyone to get paid and the studio to stay afloat so it can make more games.

      @Ffitz: I’m assuming that all of this will be played through augmented-reality-type things and not actually require you to run the block or wave a plastic pistol around.

    • Half says:

      We do have floating cars, their just fucking useless, and we could build a nuclear rocket if it weren’t for the damn conservatives worrying about irradiating the entirety of the Western Hemisphere. That’s the funny part. Futurists predictions after the dawn of modern science have all been technically true, they just like to completely forget the sociological and ergonomical factors involved in technology as well.

  4. Iucounu says:

    Good as this article is, isn’t Ray Kurzweil kinda full of shit? Nanobots in my blood projecting games directly onto my retina by 2029?

    • Jim Rossignol says:

      I’m not taking Kurzweil all that seriously.

    • Iucounu says:

      Oh, I didn’t get the vibe that you were…

    • corbie says:

      That bit reminded me of the redoubtable Kevin Warwick. Lecturer, loony and part-time cyborg. Having a garage-door-opener chip implanted under the skin your arm is apparently the future. Everyone knows this. WE ALL SECRETLY WANT TO BE CYBORGS!

      link to – its a shame the Beeb take him so seriously.

      Also what Charlie Stross says about his books not echoing the future but the present reminded me of George Orwell and 1984. People use “Orwellian” a lot when talking about the future when Orwell was talking about specfic issues in 1948.

      In the future there will be robots lunatics.

      Loved this article the first time round too Jim.

    • AbyssUK says:

      what scares me.. quantum computing… give that to a game AI.. imagine an opponent AI that already knows every single outcome before you’ve even made a move. Us simple humans have no way of beating such an AI. I demand UFO enemy unknown Quantum edition.

      Quantum entanglement communications would remove lag, so systems such as Onlive would actually work even from your spaceship.

      However sadly the future of games if it does all goes ultra realistic with full body feedback brain implants etc.. is porn and we all know it.

    • AbyssUK says:

      i have no idea why my post went here!

    • mrmud says:


      This is way beyond my understanding of physics but apparently quantum entanglement is not able to transfer useful information faster than the speed of light.

    • Tei says:

      “what scares me.. quantum computing… give that to a game AI.. imagine an opponent AI that already knows every single outcome before you’ve even made a move.”

      If that what you need, is a simple as calculate “v_playerSpeed * time = v_newPlayerPosition”, so If you want to *exactly* hit the player, make so the bot shot to v_newPlayerPosition, and It will hit with perfect accuracy. Any 4004 CPU can already beat these puny humans at this.

    • Jacques says:

      I’d take Warwick more seriously than Kurzweil. If only because Warwick isn’t so afraid about the inevitable (ie dying) that he takes a handful of vitamin pills every day.

    • Anonymous this time says:

      Warwick is an attention whore who happily accepts the credit for the work of others in his department…

  5. archonsod says:

    “Maybe we will. This much is certain: there’ll be games on that Starship, wherever it goes.”

    That nobody will be able to play, because it takes six years for the DRM to phone home and submit your DNA sample ….

    • Jim Rossignol says:

      There is no DRM or piracy in the future, of course, because the punishment for copyright violation will be to work yourself to death in the ore pits of Mars.

    • AndrewC says:

      Oh gosh I hope Mars was colonised by cockneys.

  6. AndrewC says:

    You’ll be able to get a crack from the pirate starships, invisible to the authorities by being made of a cloud swarm of tiny ships made entirely of mirrors.

  7. Sagan says:

    Kurzweil will be right about some things. Singularity by 2029 – possible. We will have the necessary hardware, the challenge will be to write software that is smart enough. I personally think it unlikely that that will happen so soon.

    Nanobots in your blood: Unlikely. Some things we will put into our blood, like improved red blood cells that make you more healthy. Robots: Never. People are too freaked out by that thought, and will remain too freaked out. Maybe a couple people will do it within a couple of decades, but it will take a much longer time before a lot of people will be comfortable with that thought.

    Related to gaming, I am completely on board with Eskil Steenberg: We will still use a mouse and keyboard in 20 years, or something very similar like a touchscreen, and games will still be very similar, except with much better AI.

    Bioware should figure out a way to enhance the Sims AI for use in RPGs, so that characters have proper motivations and so that they react to what you do without scripts. And some large developer like Blizzard should put a lot of money behind procedurally generating levels and stories within those levels. I am curious to see what Diablo 3 will bring in that regard.

    But in the long term, it’s obvious to me that Iain Banks gives the best predictions for how our technology will evolve.

    • mandrill says:

      All we need to do is write software that is capable of improving itself, then general AI will evolve naturally. Imagine an operating system that can re-write its own code for optimal performance every few seconds, always testing and improving. once that happens AI is only a very quick hop up the exponential curve from there.

    • ohnoabear says:

      Spoken like someone who has never tried to write self-modifying code. It’s very challenging, especially on a large scale.

      Heck, there aren’t even a whole lot of languages that allow a programmer to write code that changes itself. Most languages in common use don’t let you treat code as data for security reasons (which is, in most cases, a good design decision).

      Now, you can write code that performs optimizations based on current conditions in any language (in fact, most modern operating systems do this to some extent), but that’s hardly the self-programming wonder AI Kurzweil thinks we’re going to see in the next 20 years.

  8. Trip Skyway says:

    Lovely article.

  9. Monchberter says:

    The problem with progress is the cost. A billion fold in computing power is difficult to imagine, let alone implement in games. Games have a long gestation period, engines have to be built, art assets constructed, levels designed, people paid. As games get more complx, this only increases the cost, and potentially lowers the rate of return. No wonder most games released today are safe and generic?

    I think the only progress will come with procedural generation and full AI design to render and make games on the fly. Things will remain costly and small scale otherwise.

  10. symmetry says:

    Great article.

    We wouldn’t have got to where we are without the graphics. I’m sorry but getting today’s mainstream to play something that looks like (even if it doesn’t play like) space invaders would be a lost cause. Good looking stuff sells, that’s why we are a $40bn industry in the first place.

    That doesn’t mean that gaming isn’t growing up, it is. I think if games were a human, they’d be aged around 25 right now – it’s all about flashy cars and loose women. When games reach their mid-thirties to forties then we’ll start seeing some real depth.

    This is obviously a broad generalisation, you can get 25 year-olds who are very mature, and others less so.

    • SheffieldSteel says:

      I’d say games are about two. They’re starting to be troublesome, and also to exhibit some charming variation and individuality, but their accomplishments are very limited compared to adults, and the conversation frankly isn’t worth much.

  11. Jakkar says:

    My thanks for the thought-provoking read. Nothing unexpected, but nice to be reminded again of my glorious immortal eternity of gaming.

  12. Dave says:

    To me, OnLive is less likely than many of Kurzweil’s predictions. Say you have a peak load of 1 million simultaneous players online and a more typical load of, say, 300,000. How many more cores, and how much more bandwidth do you need in your server farm if you’re running both servers AND clients?

    And if computing power gets cheap enough that the cost difference doesn’t matter, all those players are already wearing a couple of hundred computers and really don’t need you to host their clients anyway.

    • SheffieldSteel says:

      All this talk of “clients” and “servers” shows your age, old timer. To the kids of the future, it will just be different boxes providing different views of the world.

  13. JUrassicPork says:

    ‘ore pits would be sweet. I’d work myself to death in days. Long live our glorious robot masters

  14. Cooper says:

    Picking up on a point halfway through. Tools, objects, things have been recognised by philosophers and sociologists as part of ‘what makes us human’ in various ways since the late 20th Century.

    What makes computers and related technology intrresting is that they are often specifically, explicitly about what we are as social creatures in the way that “the hammer” or “the jug” were not so overtly so.

    This isn’t simply mediation. It’s much more than that. As wild as transhumanists are seen to be – we are already there, already cyborgs, in many ways.

    (I use the term we in a very specific sense – those of us who can afford and have access to such technology; the importance of the ‘technological divide’ should never be downplayed or ignored)

    What’s so fascinating, so important about us cyborgs and our future is just how banal it all is, and how banal it will be – the future has happened, and it’s already become the everyday.

  15. UK_John says:

    I am a futurologist too. And within 4 years a new home entertainment system will kill AAA video gaming, leaving a small independent, download only PC market, Annual PC games sales will be at 112 million as opposed to the 12 billion it was 10 years before.

    • TeeJay says:

      You are saying that PC games had revenue of 12 billion ‘somethings’ in 2004?

      Are you going to launch into one of your cut-and-paste “we’re all doomed” posts where you pretend that annual NPD retail sales figures reflect total PC gaming revenue, that declining Gamespot scores “prove” that current games are not as good as in 1999 and that the thing that “killed” PC games was getting rid of “big boxes” and paper manuals?

      DFC Intelligence say that 2007 saw worldwide pc games sales of around $11 billion of which:

      Retail 30%
      Subs 44%
      Digital downloads 18%
      Adverts 7%

      Americas $2.7 bn
      Europe $3.5 bn
      Asia $4.5 bn

      Kind of makes your prediction of 0.1 bn sales look kind of wierd. (nb the PC has a far bigger installed base around the world than anything else and this isn’t going to change in the next 10 to 20 years – the opposite in fact).

      I have asked you before and am still waiting to see your evidence that *overall* spending on PC games is falling. From my perspective the evidence suggests a fairly steady overall level of spending with a switch first from high-street to websites and then to downloads.

  16. Mike Arthur says:

    What a load of bollocks!

    OnLive will not be replacing home computer systems any time soon, the latency and bandwidth requirements are just too high.

    People have been stating AI will be “smarter than humans” for a long time but that requires, as someone pointed out above, self-modifying code which is extraordinarily difficult to write. AI is good at crunching huge amounts of data if you give it patterns to look for and not much else.

    We used to have dumb terminals we all connected to from remote systems. They were called mainframes and most people don’t use them any more (for good reasons).

    The computer power speed increases are fairly hilarious too. We’re not seeing computers get much faster at single operations, they are just becoming more parallel and the programs are being written to take advantage of that. Clearly these futurists haven’t written much code if they thing that multithreading programming is a) applicable to all problems or b) an easy way to just keep getting better performance.

    This stuff about the singularity and the like is hilarious, people have been predicting it “coming soon” for a long time and they will continue to do so. Futurists should get out there and actually try and make some of this happen instead of making wild speculation which turns out to be bollocks.

    • Mister Adequate says:

      I’m not aware of anyone who has predicted the Singularity to be “coming soon” until fairly recent years, the last decade or two. And in fact whilst some of us Transhumanists reckon “within our lifetimes” is a plausible, if not guaranteed scenario, Kurzweil’s one of very few who sets a date.

      Before I read The Singularity Is Near I was very sceptical about these claims. But Kurzweil’s evidence is extremely strong and I was convinced by it.

      Finally I suppose it needs to be pointed out – talking and thinking about it IS trying to make it happen. It’s like saying John Locke should have got out there and made things happen instead of just philosophising about it.

  17. EBass says:

    Well I’ve been saying for some time, as graphics improve and game environments get harder and harder to build, developers will have to rely more and more on computer generated environments. I think in the next decade we’ll start talking not about what graphics engine the game is using but what liscenced algorithm the game is using to procedurally generate context.

    The comments about AI however are far premature. Sure we’re slowly getting to the point where bad guys use cover flank etc etc (although we haven’t moved on that hugely from Half Life), but we’re nowhere near the point where they can actually converse with players in dialogue. On that front we’ve hardly moved past the text AIs from the late 80s. Its not merely about processing power its coding the AI thats the hard part.

    However does anyone see a moral issue here? If we begin to code AIs that are actually relativly indistinguishable from the real thing, we might begin to gun down in hundreds computer programmes which are actually as sentient as your average small mammel.

    • Half says:

      The value of life isn’t defined solely by its processing power, nor delusional religious values like souls. The value of life is intrinsically linked to how much we value a individual of the agents variety. We live in a society that values the individual of people, because we can distinguish the diverse array of personalities, identities, and contributions of diversity. Even in a dog, we can value the individual dog because it exists as a unique entity in which we invest emotions into. Even a mice carries a value, they are spatially individuals, simply put, if I have two mice, and eat one, I only have one mouse.

      A computer program, no matter how smart, simply possess no individual irreplaceable attributes, as it has lacks relative, spatial, and social individuality. The only way to make a AI existence “meaningful” is if we allow it social individuality and the ability to relate with humans. That won’t exist outside of a lab environment.

    • Half says:

      The value of life isn’t defined solely by its processing power, nor delusional religious values like souls. The value of life is intrinsically linked to how much we value a individual of the agents variety. We live in a society that values the individual of people, because we can distinguish the diverse array of personalities, identities, and contributions of diversity. Even in a dog, we can value the individual dog because it exists as a unique entity in which we invest emotions into. Even a mice carries a value, they are spatially individuals, simply put, if I have two mice, and eat one, I only have one mouse.

      A computer program, no matter how smart, simply possess no individual irreplaceable attributes, as it has lacks relative, spatial, and social individuality. The only way to make a AI existence “meaningful” is if we allow it social individuality and the ability to relate with humans. That won’t exist outside of a lab environment. (Obviously, programs cannot have spatial individuality by definition)

  18. TeeJay says:

    Kurzweil: “Games are the harbinger of everything… [arguing that games were where the future manifested itself] Ultimately … they’re going to be competitive with real reality.”

    Sorry if I am being stupid but I read the article looking for the bit with some examples of games which ‘manifest the future’.

    Did I miss something?

  19. anonymous17 says:

    Considering that the 1960s saw the advent of nuclear powered aircraft and fusion powered space hotels predictions, Kurzweil’s predictions should be interpreted as calling forth a dead burnt-out world orbiting a dying star undergoing stellar collapse due to solar system warfare.

    Here is to the future.

  20. Jason Moyer says:

    Ray Kurzweil is awesome. I had no idea that he had any interest in gaming.

  21. Half says:


    Humans will always be Humans until their not anymore, and then it doesn’t matter. We’ll still love and die and cry and laugh and create and destroy and hate and kill and dream. Technology won’t change that. Whats all the hullabaloo about? Nothing changes :/

  22. EBass says:

    Thats bollocks half. By your argument if we could clone humans so they would not just be physically alike but have the exact same memories/thoughts/desires etc then murder would no longer be ethically wrong because now humans have no individual replaceable attributes.

    The value of life comes from its perception to percieve the reality around it, In other words consciousness, self awareness and ability to feel pain. We don’t mind slaughtering bacteria by the millions do we?

  23. WCG says:

    Wow! Great article! Is it just my imagination, or am I seeing more and more of this kind of thing here? Either way, keep it up. You’re giving me lots of food for thought (and plenty to blog about, frankly).

  24. Bong Ebbighausen says:

    Is actually blogengine as good as wordpress for some reason? Really needs to be as it’s ever more popluar recently.