Artificial Intelligence is all the rage these days and there's no shortage of rave reports (like this one) or cautionary tales (from the top of my head: hallucinations, infringement of copyright, outsized energy consumption, etc.). These are all worthy considerations but from my humble vantage point, my main take away is that, with so many interest groups involved and so much investment already committed, AI is likely here to stay. With this out of the way, I'm left craving for more information on the ways this could be put to practical use in the world of tabletop gaming, beyond generative AI.
One area of particular interest to me is playtesting: it's necessary, often data-driven and very time consuming. Game designers already use pretty advanced methods to assess their own game outcomes, like generating Monte Carlo tree search to weed out flawed or imbalanced strategies. Projects like boardgame.io try to democratize the approach by providing a development framework for it. In other instances, the whole process can get outsourced to another company as recently reported in this article from The Economist (the original inspiration for this post). This can get very nerdy real fast but my main take-aways were:
- Playtesting Automation is really helpful in quickly producing statistical data; that, in turn, is fertile ground for AI, which can learn and interpret this kind of information much faster than we do.
- It's unclear if/how getting AI involved in the design process might alter the fun factor of a game.
Have ever you used AI for playtesting, or in any other meaningful area of your board game design? Please share your experience in the comment section!
Cheers,
Ady
0 Comments