Expressive Intelligence Studio Blog

Hofstadter-inspired writing, with a helping of Knuth served in the comments. Also good in Mark's blog: Ribbon Hero is the future ( http://diveintomark.org/archives/2010/03/03/microgames ) about learning/achievement games. That post included a link to Space Invaders Enterprise Edition ( http://eis-blog.ucsc.edu/2010/02/space-invaders-enterprise-edition/ ). The title by itself is worthy, but the post is about rule engines. That interesting point was not much discussed here: http://news.ycombinator.com/item?id=1131502 .

You might enjoy answer set programming (ASP). It's a different take on logic programming based on the idea of automatically transforming your problem into a SAT-like representation and then running fancy conflict-driven constraint learning solvers on it. In many solvers, you give up Turing completeness in exchange for being really productive with solving NP-complete problems (similar to regex engines turning you into a bad-ass string matching programmer). Here's a tutorial application of ASP to map generation: http://eis-blog.ucsc.edu/2011/10/map-generation-speedrun/

Here's one experiment: http://eis-blog.ucsc.edu/2009/11/cfml-the-context-free-music-language/

They pulled the episode, so no one can (officially) see it unless they saw it last night. http://eis-blog.ucsc.edu/wp-content/uploads/2010/04/sp_1406_Sorry-500x…

No, I'm not one of those EJB guys, I'm not a fan of shoehorning technologies that don't belong to software that doesn't need it. I went to rule engines because I thought it really provided some value. I elaborate on this on a comment back at the blog http://eis-blog.ucsc.edu/2010/02/space-invaders-enterprise-edition/com… DISCLAIMER: I'm the author of that post.

I'm using it for modeling videogame prototypes in logic, and then deriving properties of them, aimed at providing richer design support: http://www.kmjn.org/publications/Playtesting_AIIDE09-abstract.html The longer-term goal is something CAD-like, where you get immediate feedback on design changes, analogous to how architectural CAD systems will highlight things like "lacking structural support" or "violates building code". With games simple versions could be "unreachable areas", "item never needed", etc. More involved versions could be generation of basically the kind of log data you would get from playtesting, only instead of empirical log data, it's analytically generated logs of possible playthroughs exhibiting requested properties: http://www.kmjn.org/notes/analytical_metrics.html Many of the above can also be done with software-verification systems, which in some cases might be a better choice for efficiency reasons. However, the stuff coming out of the AI community is a lot more flexible as a modeling language, feels less like writing a fixed specification. It's also designed to be editable, e.g. you can add or remove game mechanics without re-formalizing the whole domain (a property John McCarthy calls "elaboration tolerance"), due to being based on a nonmonotonic logic. I was also already familiar with Prolog, and ASP's syntax is heavily modeled on Prolog's, even though they work entirely differently under the hood. As far as getting started with it, the Potassco…