I was at AI IDE 2008 today – an AI conference at Stanford. I’ve got notes that I’m hoping to publish, after I get it cleared through all the channels. Long story short, general AI research is getting a new life now that procedural is the new buzzword, and there’s definitely a great industry focus and a broadening of research accordingly. There were 2 specific industry driven talks, and one that even specifically dealt with tools! But there’s also still a lot of algorithms being proposed for advanced AI that are impractical for 95% of the games out there, and a general avoidance of game design importance. Academics are catching up, and the push in college programs towards dedicated games programs and games professors is going to pay off for the entire industry.
I was talking with Pat Olivier (I believe) and the proposal came up, as it is wont to do in AI circles – “How long will it be before AIs are replacing our humans in games?” The designer and the AI guy in me clashed. There was a circle of fire, spikes, chanting, side betting, the whole works. But looking at market data always wakes me up. The data just doesn’t support this. We use AIs now to communicate experiences that we couldn’t do any other way. But the reality is, given a choice, humans prefer to play against other humans. World of Warcraft, Counter-strike, even GTA, Madden, these are games people play with other people. Far and away. It’s a similar problem to what David Cope spoke about today – there is little joy in playing against a machine that you know will exist, repeatably, forever. Chess is “solved”, people train against computers at chess, but far far more people play chess against other humans.
The counter argument is that yes, this is true now, but what if the AI passes the Turing test?
Well, assuming such a thing is reasonable over the long scale, even then I don’t think it will matter. If you could have a meaningful relationship with an AI you would by definition be engaging in it. But would you only engage with AIs of that type? I don’t think there’s much evidence of that. People engage with their family, their co-workers, their spouses, their lifelong friends. Maybe some people engage with their dogs more then they do with other humans, but that’s a rarity. It seems likely that if AIs do start impersonating these other humans, people will just demand tests that differentiate. Because it’s no fun to be friends with someone who deep down, you know, is just telling you you’re right because they want to be your friend.
But what if they aren’t just game entertainers or companions, they have intrinsic human motivations that don’t rely on consumers? It’s allows possibly to say but what if they are just more human in that way. But each one of these depths is a new scale of learning, and we still aren’t even baby stepping down the path. I think a better question would be: Will humans ultimately create such an AI? Or will it be an AI creating this AI?