Congratulations! You’ve completed level two of the GDC! Here are the secrets you discovered:
- You found a novel approach to modeling social situations in a computer
- You discovered that you are not the only one who thinks old-school video game soundtracks were actually really carefully composed and that some are absolutely f***ing brilliant, on par with compositions by Bach and Mozart
- You discovered that Phil Fish has really strange thoughts about games, too
- You rationalized a beggar
The first talk of the day was a panel of AI researchers and video game designers on the topic of how to use AI techniques to meaningfully model interpersonal relationships in a character-heavy game. For the most part, the techniques they covered seemed to be fairly typical “connected graphs with benefits” approaches. This has the advantage of being intuitive and obvious, but suffers from the disadvantage of being incredibly complicated and often quite brittle. One the panelists in talking about his game “Prom Week” mentioned the idea of a “social physics” model. This is a great idea, but in practice what they had done was try to model a social landscape from the ground up, listing out all the possible interactions with and attitudes toward other characters that each character could have. It’s like trying to simulate physics by listing out every possible reaction to other particles that every particle could have, in a giant switch. Without an overarching model or theory to tie all these reactions together into a few simple equations, you’re inevitably going to miss some important detail. The last speaker in the presentation, however, talked about a model he’s using in his game that very well could provide such a structure. It makes use of the “big five” personality traits from psychology to model the fundamental abstractions behind the social states people are in as a kind of “social DNA”–each state, emotion, affect or quality is expressed as a 35-vector of minuses and plusses to each of the big five, as well as to each trait’s six sub-traits. This system offers the advantage of being totally agnostic in terms of actions, beliefs, and dialogue, all of which can be layered on top of and tailored to be specific to the system. It also offers the advantage of being able to apply these traits to inanimate objects in the game–for instance, a particular location might have traits that increase sociability and curiosity, or a particular object might be associated with increased courage. The disadvantage, of course, is that it still leaves the task of “sequencing” the social genome to a 35-vector with potentially endless combinations of values. I noticed that the speaker quietly left that part as “an exercise to the reader”. If it proves viable, however, it could provide a tremendous opportunity to create far deeper and more expressive interpersonal interactions than are currently possible in video games.
You discovered that you are not the only one who thinks old-school video game soundtracks were actually really carefully composed and that some are absolutely f***ing brilliant, on par with compositions by Bach and Mozart:
This chick also thinks so.
In his soapbox talk on Fez and the nature of the video game medium, Phil Fish was pretty vague. He talked about comic books and four-dimensional cubes and the possibilities of choice, and a lot of it didn’t make much sense. But it resonated with me anyway, much like Fez resonates with me deeply despite the fact that I have not yet actually played it. I’m looking forward to Fez, and I’d also like to meet Phil personally and pick his brain a little, see where we agree and where we disagree on the subject of games as a medium. So, the next time I see him I’m going to introduce myself, strike up a conversation, and see if it goes to an interesting place. Wish me luck!
After the conference, as I tried to avoid eye contact with what was probably the seventh or eighth homeless beggar I had seen that afternoon alone, I caught myself actually thinking “you know, he probably doesn’t deserve my money anyway; I bet he’d just waste it on booze or something”. As a rule, I try to avoid these kinds of judgements, because I strongly believe that most modern human problems have their fundamental cause in lack of empathy. Judging from this experience, it seems that the more exposure you have to the homeless and impoverished begging on the streets, the less likely you are to empathize with them. Seems counter-intuitive–until you think abut the cognitive dissonance involved. If you only see one homeless person a week, you can give them money every time you see them; no dissonance. If you see two an hour, however, you can’t possibly give all of them money, and so you start ignoring them. Now it’s actually harder to give a homeless beggar your money once a week than when there was only one, because it’s an exception to your default behavior, and defaults have a lot of psychological momentum. Often, you’ll stop giving altogether. Since it’s considered good to give money to the homeless and you don’t, and you consider yourself a good person, you have dissonance to resolve: hence the rationalizations that the homeless are lazy, alcoholic, that they’ll waste your money on drugs and booze, that they could make a living and get off the streets if they only tried, etc. Put in this light, a city’s incentive to rid its streets of homeless (ideally by, you know, giving them homes) is twofold: it helps out the people living on the street, but it also makes the people living there who do have homes more likely to empathize with the homeless, and hence more likely to support programs for helping the homeless get off the streets. It’s a positive feedback loop.
I always have dark thoughts about the homeless when I’m in San Francisco, as it’s a particularly bad problem here. In a future (non-GDC) post, I’ll go into more detail about the issue, and about the problems I see inherent in anti-welfare positions. For now, I’ll only say to be careful of the rationalizations you’re making. As I saw today, they can all too easily start making you into a person that you don’t want to become, all without your even being conscious of the change.
You’ve unlocked level three! Continue? (Y/N)