When games are Science

Posted by AlanZucconi (twitter: @AlanZucconi)

January 2nd, 2014 9:09 am

No, no, this is not going to be the same boring, old discussion about games and art. Games are Science, full stop. Douglas Heaven has published today a very interesting article on New Scientist about Michael Cook entry.  Or, to be more precise, the entry of ANGELINA, the game-developing AI he’s currently designing at Imperial College London / Goldsmith University. The article features the following games: show them your love for Science!

 

To That Sect by Michael Cook / ANGELINA

 

0RBITALIS by Alan Zucconi

 

Cat Gentlemans Play: Insult Spinner 10 Cents by RobotLovesKitty

8

This entry was posted on Thursday, January 2nd, 2014 at 9:09 am and is filed under LD #28. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.

5 Responses to “When games are Science”

1. mtrc says:

   January 2, 2014 at 9:46 am

   Thanks for writing this up, Alan! I’m excited to see the final results for the game.

   Log in to Reply
   • AlanZucconi says:

     January 2, 2014 at 10:24 am

     Regardless the final results, ANGELINA got a lot of attention! And I am pretty sure that an AI making a game for LD was a first anyway! 😀

     Oh, speaking of firsts… :p …DO COME TO LONDON INDIES! 😀

     Log in to Reply
2. goerp says:

   January 3, 2014 at 4:52 am

   Hmm, I’m not sure if the ‘Games are Science’ discussion is more interesting then a ‘Games and art discussion’. In any case there are a few things about this post (and the New Scientist article) that rub me the wrong way.

   There is of course the problem of defining what creativity really is, but regardless, I think it’s pretty hard to create a really creative result. To have a result that is more than recognizing emerging patterns in games we allready have.
   I am very curious about where this will go, but for the Ludum Dare I think this is terrible.
   Why people do the Ludum Dare varies (doing something they haven’t done before and trying to impress fellow game makers, just getting better or maybe even getting into the business through this) but giving them the message that computers can do it better than they is just bad.

   As an one time experiment I can see the point (a sort of Turing test), but planning to enter this regularly I think, is insulting to game makers. ‘I made a program that creates better games than you’.

   Log in to Reply
3. sorceress says:

   January 3, 2014 at 6:19 am

   As an one time experiment I can see the point (a sort of Turing test), but planning to enter this regularly I think, is insulting to game makers.

   Since when has ethics stood in the way of scientific progress? 😉

   I made a program that creates better games than you

   Well this would be a bit deceptive in my opinion. Afaik, the generator uses pre-written game templates, and merely selects assets and parameters for it. So there is very much a human ‘creator’ of the game(s), who has no doubt spent months working on these game templates. So if the games are “better”, then the vast majority of that credit goes to the human who created these templates, rather than the AI itself.

   The whole process is very similar to making a collection of games that have procedurally generated levels. But rather than the procedural generator selecting assets and parameters randomly, it performs a tag search based on word associations with the theme. And we’re calling it an AI instead of a procedural generator.

   I’m not sure if the ‘Games are Science’ discussion is more interesting then a ‘Games and art discussion’.

   I’d say that games are both an art and a science, just like cooking, and music. They’re scientific in the sense that you need the game/cooking/music to work, and that takes study and experimentation, and training. It’s easy to create broken software, inedible food, and painful noises. In ignorance, it’s much easier to fail than it is to succeed.

   Games are artistic in the sense that “stuff that works” isn’t a single point that we must focus in on, but a vast space that we have creative freedom to explore. The better we know and feel that space, the more confidently we can operate within it, and enjoy personal variation.

   Log in to Reply
4. goerp says:

   January 3, 2014 at 6:44 am

   Maybe my response was a little too heavy handed. The message of ‘it makes games better than some human entries’ came from another game maker who was interviewed not by the researcher.

   I don’t want to turn this into a rant so I’ll try to keep an open mind about this.

   I’m pretty sure that in the end you can simulate creativity and ‘fool’ people in thinking it was made by a human. I’m not sure if it’s the same as human creativity.

   Log in to Reply