New paper on procedural animation

This is a longer, journal paper on the Twig procedural animation system, submitted to the new journal IEEE Transactions on Computational Intelligence and AI in Games (full disclosure: I’m an associate editor).  This paper documents a fair amount more than the AIIDE paper, including object manipulation scripting, and a little bit about the authoring tools.

New paper on inference/action-selection – comments welcome

This is a paper I wrote for ICFDG that takes some work from my previous life as a roboticist and applies it to Twig (the procedural animation system I’ve been working on).  However, it’s situated in a tradition that the games community, and even the present-day AI community isn’t very familiar with, so much of the paper is an attempt to survey and explain the literature on using combinational logic for inference – why you’d want to do it, why it’s limited, and what you can do to overcome some of the limitations.  And even then, I’m worried that it won’t be transparent to the reader; not because it’s technically difficult, but just because the motivations won’t be clear.  Anyhow, any advice would be welcome.

fall symposium talk slides

Here are the slides from my position paper at the AAAI Fall Symposium on Naturally Inspired Artificial Intelligence.

my other blog

I’ve been maintaining a separate blog that’s specifically for Twig, the procedural animation system I’ve talked about a little here.  You can find the blog here.  At the moment, it’s more active than this site because that’s where most of my limited hacking and thinking cycles are going.

The blog also includes pointers to the episodes of a “webcomic” that I’m doing with Twig as a way of testing out features and guiding feature implementation.

Final version of AIIDE paper on procedural animation

This is the revised version of the AIIDE paper on Twig, the procedural animation library I’ve been working on.  I’m going to do a longer journal version of it, so suggestions are very welcome.

Another paper

Men are Dogs (and Women too).  This one’s for the AAAI Fall Symposium on Natural Computation.  It’s my latest effort to try to get my intuitions about cognitive architecture onto paper.  There are a bunch of other arguments that I want to fold into it, but it’s already overdue, so they’ll have to wait for the camera ready copy.  Comments and suggestions welcome.

Procedural animation paper

Sorry for being off the air for so long.  This is a paper I did for AIIDE on the procedural animation work I’ve been doing.  Unfortunately, rather too much of it was written during a 39 hour plane-trip-gone-bad, so I was pretty seriously sleep deprived when I wrote it.  Comments welcome.

Things are winding down for me at school now, so I should be posting more.  There’s another AI paper coming in a few days.

position paper: ai and psychopathology

I’ve just put up the final version of my position paper for the AAAI Symposium on Intelligent Narrative Technologies.  It argues that popular narratives deal at least as much with the ways characters depart from ideal rationality as it deals with actual rational behavior, that we don’t have computational theories of those modes of behavior, and that we could learn a lot by trying to model them.

make love, not war

I was working yesterday on adding fighting behavior to my procedural character animation.  In particular, I wanted the kids to be able to shove one another around.  But due to a bug they kept approaching one another rather than stopping at “shoving distance.”  So they grapple in what looks like a particularly comical make-out session.  Possibly useful, but best saved for the parent characters…

believability and attributional intentionality

Robot hackers have long realized that human observers tend to over-attribute intelligence, or at least intentionality, to robots, at least provided that they’re moving.  (Dave) Miller’s law states that the perceived intelligence of a robot is directly proportional to its velocity (Dave didn’t name this Miller’s law, but he said it once at a workshop and I’m fond of quoting it).

The image above is a screenshot of what appears to be two child characters playing with one another while being watched by an adult.  However, in actuality, what I’ve been implementing is attachment behavior, which is the response of children to stress by seeking out their caregiver (more on this another day).  To implement that, I need to have something to stress the kids out.  The right way to do it is to implement a real social engagement system with wariness and coy behaviors, play, turn-taking, etc.  However, the first step in that is simply to make a second child and then hack the children’s appraisal systems to assign negative valence to strangers (i.e. to each other).  All that does is make the kids watch one another and keep their distance from one another.  For example, one won’t approach the ball if the other is too close to it.  There’s no real sociality going on there.

The interesting thing is that it’s enough to make them look like they’re playing.  They both run to the ball, but then when one gets to close to it, the other backs off.  The first one will kick it until it happens to kick it toward the other one (which is pretty frequent since I haven’t implemented aiming).  At that point, the first one stays away from the ball and the second one plays with it.  This continues until they get far enough from the parent to engage the attachment system, at which point the attached child runs to the parent and hugs him/her, then runs back to play.

The point of this isn’t that this is a good simulation of anything, just that surprisingly simple behavior can appear engaging and intelligent, provided that whatever behavior you do have is relatively fluent.