Ragdoll Metaphysics: Milo and Kate, or Artificial People Are The New Games
Microsoft's motion recognition system, Project Natal, presents an abundance of possibilities for the future of videogames. The discussions arising from the abolition of the physical control device were one of the highlights of E3: a chance to see entirely new game mechanisms, and to break down the barrier that seems to have been created by complex interfaces.
But it's the appearance of Milo - a simulated, on-screen boy who responds to human interaction - which is perhaps the most telling event. For sheer novelty alone he stands out: he is a modern day automaton, a simulation that we can't help looking on as a little bit alive.
This specific direction of Natal tech also suggests a number of futures for videogame evolution. It's moving towards the interface that human beings most readily understand: language and interaction with other people.
What's critical about Milo's unveiling is that he's tied to a new technology which allows a game console to recognise users, and respond to physical cues - movement, speech, even physiognomy. The artificial person as a character in our everyday gaming lives is suddenly just a little bit more concrete.
Of course no one believes that Milo is anything more than set of call-and-response tricks, with little flexibility, but both he and the system used to bring him to life point the way to much more. Imagine where this could go: a computer intelligence that shares your gaming with you, or you music with you, or your Googling with you.
Perhaps it could be a character that "lives" in the menu screen, showing you new things, suggesting new experience you might enjoy, or helping you search for content. Or perhaps it would go deeper than that. Just as Mii avatars have been able to cross-pollinate their way into games, so these characters might accompany us into fantasy worlds, or science fiction settings.
Perhaps game engine development might become a secondary concern to character development: the research and development that makes up games could come rely on creation of interesting artificial people, rather than new, stunning visuals, or extra technical trickery. The challenge becomes one not of simulating worlds in more detail, but simulating humanity with greater fidelity. The gaming race would become about building someone interesting enough to keep players engaged, and to keep them coming back.
Of course, the appearance of Milo is not the first time a simulated person has wowed us in the world of videogames, because the evolution of artificial people has been racing along on a number of fronts. The most obvious of these is the artfully crafted non-player character - the Alyx or Barney of Half-Life 2, or the crew of Mass Effect's sleek starship. These clever mannequins have been structuring our game experiences for years now, and they are only become more believable with each passing year.
Another important strand in the development of the artificial person is represented by the bots who can fill in for human input in multiplayer games. The actions that we need to perform to play, say, Left 4 Dead, are now readily executed by programmed routines. These are limited, separate aspects of human activity, but they are starting to add up to a whole. The only thing our artificial people can't quite seem to manage is to type comprehensible, dynamic dialogue. That might soon change.
Returning to the example of Half-Life 2's Alyx, we can already see that building even a rudimentary kind of relationship with a character is becoming crucial to game design. You might not be able to talk to Alyx, or even genuinely interact in any way, but she talks to you, and that charm means you respond more readily to the emotional cues in the game. This, already, demonstrates the value of an artificial relationship within a game world. Imagine the impact when Alyx can really "see" you, and respond to your smile, or your rude gesture.
Another strand of the artificial person is the social behaviour of our favourite pocket soap opera, The Sims. While these are, again, only rough analogues of real people, they demonstrate something profound about how we are disposed to react to people-like creations: if it behaves a bit like a person, then we're not only engrossed by its activities, but we also behave towards it as if it really is alive. Imagine the possibilities for future Sims games where the emotional and behavioural fidelity is that much greater, or in which Sims look out of the screen and respond to you directly.
All these trends feed a single idea about the future of games: that it will be furnished not simply with new challenges and experiences, but with new relationships. Artificial relationships. Legendarily verbose games journalist Tim Rogers made a couple of astute observations about Milo: the first being that if the technology were used to create a pretty girl, Microsoft might finally crack Japan... [ed. note: something that Microsoft did in fact flirt with on the original Xbox with its import-only N.U.D.E. (pictured)]
Less sleazily, the possibilities for game design are thrown into relief, suddenly giving us access to conceits that were unimaginable. The obvious game, says Rogers, would involve us being Milo's imaginary friend. "Who are you talking to?" ask the characters in the game world. "Oh, no one," says Milo. Only we are there, and he "knows" it.
Of course it's hard to believe that gamers will accept Milo unless he's a space marine covered in orc gore ("THE EMPEROR APPROVES OF YOUR PICTURE. NOW PUSH ME ON THIS SWING"). But nevertheless there's every reason to believe that the craft of the game designer will soon encompass the same territory currently inhabited by novelists, dramatists, and actors: in creating believable people for us to respond to, and to empathise with.
After all, games have subsumed every other medium, from music to firework displays, so why shouldn't the more nuanced aspects of human art and culture be any different? And of course games will do it better: just as the rich taste of interaction makes action games a superior experience to action movies, so artificial people and our relationships with them will make videogames more emotionally involving, more upsetting, and more rewarding than any novel of human melodrama or movie about the lives of ordinary people, has ever been.
The future of games, I'm willing to wager, is artificial life.
[Jim Rossignol is an editor at RockPaperShotgun.com and the author of This Gaming Life, an account of the life of modern videogames and some of the people who play them. Ragdoll Metaphysics is his Offworld column exploring and analyzing gaming's vast world of esoterica.]
9 Comments
Leave a comment
AT |
|
Review: A week with the Sumosac Gigantor
Jewelry that spells out its function, literally
Coffee table with a built-in firepit
ComiCon: Day 1 [Verdict: Nerdywood!]
Recently on Offworld: Left 4 Dead again, Catan on iPhone, One Man Rock Band
Pittsburgh Dpt. of Zombie Disposal
Netbook hackintosh chart updated
savetherobot
#1 – 11:40 AM June 10, 2009
Alyx is a great example. As Michael Abbott pointed out recently, the folks at Half-Life put an inordinate amount of time into developing that character, and she already is slightly responsive to your actions: if you flash your flashlight in her face, she puts her hand up and looks annoyed. It's a small thing but it's incredibly convincing.
Rachael Webster has written about this as well, looking at the idea of Matt Hazard as a character who exists outside of the videogames he stars in, or questioning whether it's a creative error to pose videogame heroines in Playboy, since it's essentially forcing them to act out of character. (And of course, Webster is a fictional character herself, who happens to write a real game blog.)
RedShirt77
#2 – 2:27 PM June 10, 2009
I think there are a number of ways gaming can integrate this technology. As I said the other day, a chose your own adventure style mystery game where a character travels with you throughout the game, or perhaps a bot that manages your system and then can play with you in certain games and can take some direction in a Multi-player online game. Or something as simple as interactive characters that can engage with you anywhere along the way in other more traditional games.
I think though that the future of technology is in exactly where science fiction puts it. as the personality of your home management system(hello computer), or replacing the card catalog or similar system at the library(like in time machine) or calming you down from your dashboard as the voice of your navigation system and also doing all types of things, like yoga or other instruction.
I do think that the system will be far to glitchy for some time to come for it to really be something a hardcore gamer wants to actually control a game with. I get annoyed when there is a slight delay between button smash and gun shooting so I can only imagine how quickly I would turn off a game that was waiting to process an image of me moving to see what I want to do.
AirPillo
#3 – 4:41 PM June 10, 2009
Caveat: "...if it works"
kurai
#4 – 11:56 PM June 10, 2009
You are forgetting Sega Dreamcast's Seaman, which was one of the first convincing example of artificial life.
overunger
#5 – 2:28 AM June 11, 2009
Yeah that's all great and everything-- When can we get Alyx walking around my house in beach wear?
These are important things here!
bishely
#6 – 5:29 AM June 11, 2009
Two thoughts:
One, if "game engine development might become a secondary concern to character development: the research and development that makes up games could come [to] rely on creation of interesting artificial people", then game development times would balloon disproportionately to the added depth of experience gamers would get - sure, Alyx sounds diverting, but writing enough dialogue questions and responses (to make the experience realistic enough to avoid old-school text-adventure "I do not understand "suck"" pitfalls) would take years, and wouldn't really make the game any more compulsive... That is, assuming we're not talking about developing REAL AI.
Even then, it's moot how much of a particular AI engine would be transferrable to other projects - it'd likely be necessary to write thousands of pages of script for every character for every game, before you even got down to programming what they should say and when.
Two, this bit:
"it could be a character that "lives" in the menu screen, showing you new things, suggesting new experience you might enjoy, or helping you search for content."
Sounds an awful lot like the MS Office paperclip to me, which was a great novelty for about five minutes (and maybe another couple of minutes while you checked the other 'helpers') but quickly became a major annoyance.
The risk with making things *seem* more human and approachable is that the moment the illusion of sentience is broken (whether by a misunderstood command or verbatim repetition of previously-heard script) the end user becomes far more irritated than they would if the illusion hadn't been established in the first place (compare and contrast those phone systems which present you with a menu and ask you to press the relevant number with those that ask you to say what you want and respond with "I didn't catch that, try again" over and over).
Don't get me wrong - I'd love to see AI being pushed by the gaming sector - but for that to happen, development budgets, team sizes and programming hours will have to skyrocket.
savetherobot
#7 – 9:02 AM June 11, 2009
I completely agree that artificial life will be crucial to games. But Rossignol is right to focus on artificial life and not artificial intelligence. Or 3-D animations. Text works too. Think of all the fictional characters on Twitter, or the romance plots in RPGs, which are largely text-driven - and entirely scripted.
People like to form believable relationships with fictional characters. That's been proven. But the players don't care how it's implemented.
Slurpee
#8 – 9:43 AM June 11, 2009
I can't wait for this tech to be used with a Pokemon game. I choose you Milo. :P
chmmr
#9 – 12:06 PM June 13, 2009
I'm not really seeing the kind of analysis and skepticism I'd like out of the journalism surrounding Natal, and Milo specifically. It's you guys' job to not only cover new developments, but to think one step ahead of the hype and figure out what the future might actually hold. Thus far people seem to take Molyneux's promises, and his thesis as to its ultimate meaning for games, at face value.
It's pretty clear that gamers and humans in general (partially overlapping subset?) respond really strongly to the *idea* of an artificial human. Understandably enough; it appeals to our narcissism and social reasoning faculties. In this general sense, character-driven games are a massively underexplored area in game design. I want to see more of it.
But of all the examples that Jim cites, what have these characters been good for? Often, they're used fairly cynically by developers to extract emotion from their audience. Their implied interactivity far outstrips their actual interactivity. An enormous majority of the work in character tech has gone into the pure appearance of the characters rather than their simulational underpinnings, and we're still deep in the crevasse of the Uncanny Valley even in the visuals department - and if you don't think we are, you're probably standing in it. This is a real problem for an interactive medium; it means that we're building on a foundation of sand.
The thing about illusions, such as the one that Molyneux at least owns up to being such, is that they fail ungracefully in the presence of interactivity. We might reach the point soon where humans in CG films are 100% believable - maybe - but here in videogames we still have miles and miles to go, and the nature of this road is that we can't even see clearly what the benefits and potential applications are at the end of it. We are confusing means with ends.
I don't buy characters in modern games. I see right through them. I feel like they're aimed purely at making me feel something, which is the most crude and instrumental form of artistic expression there is. It troubles me to no end that "making people cry" is considered the highest feat to which many developers aspire, and that shoving mannequins in our faces is how we've decided we're going to achieve that.
We must hold games to a higher standard. Calling the current state of the art "good enough" is only going to screw us or make us look very, very shortsighted and silly later on.