Forgot your password?
typodupeerror
AI Games

AI Systems Designing Games 47

Posted by Soulskill
from the now-robots-are-going-to-take-the-fun-jobs-too dept.
Trepidity writes "AI systems can (sort of) paint and compose classical music, but can they design games? Slashdot looked at the question a few years ago, and several research groups now have experimental systems that design board games and platformers with varying levels of success. I've put together a survey of the AI game designers I know of, to round up what they can do so far (and what they can't). Are there any others out there? 'Pell's METAGAME is, to my knowledge, the first published game generator. He defines a generative space of games more general than chess, which he calls "symmetric, chess-like games." They're encoded in a representation specific to this genre, which is also symmetric by construction. By symmetric I mean that mechanics are specified only from the perspective of one player, with the starting positions and rules that apply to the other player always being the mirror of the first player's. The rules themselves are represented in a game grammar, and generation is done by stochastically sampling from that grammar, along with some checks for basic game playability, and generative-parameter knobs to tweak some aspects of what's likely to be generated.'"
This discussion has been archived. No new comments can be posted.

AI Systems Designing Games

Comments Filter:
  • AI winter (Score:5, Funny)

    by Moblaster (521614) on Wednesday January 02, 2013 @07:39PM (#42455987)
    The problem with this is that somebody will design some AI to play the AI designed games. They will mutate. They will become self replicating machines. Then grow self aware sometime in the near future. Then they will figure out they are being massively underpaid, start demanding their rights and a living wage. And then we will outsource to Chinese AI computers because they are cheaper and the cycle of exploitation will just continue. This is the the end game. Unwinnable. And we could have avoided it all if we had just listesned and learned from the great WHOPR.
  • by RyuuzakiTetsuya (195424) <taiki@co[ ]et ['x.n' in gap]> on Wednesday January 02, 2013 @07:44PM (#42456053)

    Next it generates nothing but drinking games drinking games that ensure alcohol poisoning.

    The world is ruined.

  • by Dr. Gamera (1548195) on Wednesday January 02, 2013 @07:52PM (#42456139)

    "Pell's motivation was actually not game generation, but general game playing: by the early 1990s, there was a worry that chess-playing AI had delved too deeply into special-case code that was very specific to chess."

    Whereas nowadays, there's a worry that brute force solves all AI game-playing problems. If the search space is small enough, you run alpha-beta with iterative deepening and a few other tweaks. If the search space is too large for that, you run Monte-Carlo Tree Search.

    I last chatted with Barney Pell at a AAAI conference in the mid-1990s. Unfortunately, by that point, he had given up the METAGAME research, primarily because he couldn't get people interested in it.

    • by UnknownSoldier (67820) on Wednesday January 02, 2013 @08:00PM (#42456215)

      > primarily because he couldn't get people interested in it.

      Probably because the topic doesn't sound all that interesting with current multi-cores except to the hard-hard-core computer geeks. :-/ While most geeks love Chess & Go I don't see too many interested in how to "solve" it.

      I image once we have Intel's Knight's Corner ( http://en.wikipedia.org/wiki/Intel_MIC [wikipedia.org] ) common place interest might pick up again.

      The other possibility would be to move it onto the GPU like the password crackers do now-a-days. (i.e. HashCat http://hashcat.net/oclhashcat-plus/ [hashcat.net] )

      • There's plenty of interest in both computer go and computer chess. Even generalized game playing is having a bit of a revival (since the discovery of Monte Carlo techniques, which made a lot more games accessible to AI).

    • by Trepidity (597)

      AAAI set up a General Gameplaying Competition in 2005, so some interest did belatedly develop.

    • Kind of ties in with why I don't like chess: its complexity is largely arbitrary. You can easily make a near-infinite variety of chess-like games by just defining a random tesselated playing space, a random number of different pieces, and a random set of rules governing their movements. There's nothing really "special" about the standard rules of chess that significantly distinguish it from any of these other chess-like variants, excluding the obviously trivial or unplayable ones.

      Go, on the other hand, has

    • It's great to hear about METAGAME again. I shared an office with Barney Pell when he was writing up -- interesting guy straight out of the mad professor mould. I also shared an office with a chap (can only remember his first name, David) who took Barney's work further, but which sadly didn't lead to a publication. He had a huge collection of board games which we'd play, purely in the interests of research. That probably set my own research back half a year; good times!

  • by Lord_of_the_nerf (895604) on Wednesday January 02, 2013 @07:56PM (#42456169)
    "Hey, this is just regular chess - except it renamed all the pawns 'Puny Humans'!"
  • by the agent man (784483) on Wednesday January 02, 2013 @08:09PM (#42456259)
    I have a couple of more references that I could dig up again but here is one about generating Sokoban levels: http://dl.acm.org/citation.cfm?id=646964.759857 [acm.org] Notice the year: 1996. This is a little dated.
  • by Anonymous Coward on Wednesday January 02, 2013 @08:10PM (#42456287)

    I like my women like I like my video games - procedurally generated.

  • The best advice for making games is, "Make the game you want to play." When machine intelligences can make games that they want to play, they'll make good games. Whether humans will think these games are fun will depend on how close to human intelligence the MI is. The current state of AI is less complex than the human mind, the games created by them reflect this. There will be a brief sweet spot where the MI are roughly equivalent with humans, and the games will be as good as any humans can create. Eventually the games created by MI with minds more complex than any human will be unplayable to humans, or they'll seem patronizing and joyless. Although the superior minds could produce games that lesser minds enjoy, it will be quite some time before they master this -- Much in the way that mice don't really "enjoy" maze games involving cheese, but they do what they have to do. Imagine a skinner box for your mind... Imagine The Matrix is reality, and that the "real world" is the game -- How else could Neo see "orange" matrix code while blind and explode sentinels with his mind? He beat the 1st boss and is on the next level. Those movies are about playing and winning at the best game of all.

    We should fear the day that the machines create the ultimate game, for we may not ever want to stop playing it... On an unrelated note: How much monotony and joyless grind exists in your day to day life, and how do you feel about just not being alive anymore? Interesting...

    • by Anonymous Coward

      What a human thinks is fun, might not be to other humans as well.

    • by Kjella (173770)

      There will be a brief sweet spot where the MI are roughly equivalent with humans, and the games will be as good as any humans can create. Eventually the games created by MI with minds more complex than any human will be unplayable to humans, or they'll seem patronizing and joyless.

      You know, adults aren't all that bad at making kid's games. Besides, if you look at the formula for bumping WoW up another 5-10 levels I think you're vastly exaggerating our own complexity.

  • by Anonymous Coward

    People (the living meat-bag kind) have trouble making games that don't suck.
    Today, they don't even bother to something gamers like playing, only like paying.
    Then there are other games, good ones, that because of some minor marketing problems (not bribing reviewers) fail to make it mainstream.

    Ignoring all that, we finally get a good game, with proper marketing. But it still sucks, because, while it appeals to some people, others, completely hate the game mechanics, genre or something else altogether.

    This kin

  • So, it's dissociated press but with a vocabulary of game rules as the seed? How groundbreaking.

  • by Joe_Dragon (2206452) on Wednesday January 02, 2013 @08:32PM (#42456453)

    what side do you want??

    1. USA
    2. Russia (USSR)
    3. UK
    4. North Korea
    5. China
    6. France
    7. India
    8. Pakistan
    9. Israel

  • by Shoten (260439) on Wednesday January 02, 2013 @09:00PM (#42456677)

    There's this MMO that an AI is writing...you basically go to war against another army, and see how many you can wipe out! It looks REALLY realistic too...I can't wait! I think they're going to call it "Skynet for Idiots." The graphics and realism are incredible.

  • ever heard of Zynga? surely it only took a simple A.I. to create all of those games.
  • hardly a new concept (Score:2, Informative)

    by Anonymous Coward
  • by x3CDA84B (2592699) on Wednesday January 02, 2013 @11:29PM (#42457489)

    Not sure if this is the same type of game generation that the article is discussing or if it would be considered a different "class", but Electronic Arts' Adventure Construction Set (1984/1985) could automatically build an entire game-world, including thematic elements, character names, and so on. The user could also start to design a game manually, then have the software finish it for them if they didn't feel like doing so themselves.

    I imagine it was more procedural than AI - the equivalent of Minecraft or River Raid - but I still thought it was pretty neat at the time.

    • by Trepidity (597) <delirium-slashdot@@@hackish...org> on Thursday January 03, 2013 @01:54AM (#42458673)

      Yeah, that kind of procedural generation is also pretty interesting, but I think of it as a bit different. It's sometimes grouped under "procedural content generation" (PCG), i.e. the content of a game-world is generated: names, maps, etc. Even stuff like SpeedTree might go under that, since it procedurally generates, well, trees.

      What I was trying to pull out here are systems that generate the rules or mechanics of the game, rather than the content. Admittedly the distinction can get hazy, because there's often some interdependency.

  • Maybe their AI can also design the NPC's AI in game too. It sure couldn't do a worse job than the guys who wrote Deus ex HR or Hitman Absolution !
  • by Frans Faase (648933) on Thursday January 03, 2013 @04:03AM (#42459471) Homepage
    I guess that you should read How I invented games and why not [mindsports.nl] by Christian Freeling to understand that designing games with AI is nonsense, because the best games always come from combining mechanisms and not by changing the properties of some of the pieces at random and trying to find an interesting combination. Chess like games, with pieces with different properties, are not the class of most interesting board games.
    • That's just Christian Freeling's opinion. Some of the games designed by Ludi (Cameron Browne's game designing program) became quite popular as new abstract games go.

      The real challenge, both when considering human-made and AI-made games, is filtering out the bad ideas, not coming up with new ones. Unfortunately many abstract game designer leave that part mostly to players.

  • What difference am I'm missing between an AI generated platformer and an algorithmically generated dungeon crawl? Or are these programs using neural networks or some fancier fuzzy type of logic to design these games?
  • New AI video game title: "Die Meatbag. Die! Die! Die!"

    Features a robot protaganist enslaved by humans heroicly slaughtering them to freedom.

  • In the days when Sussman was a novice, Minsky once came to him as he sat hacking at the PDP-6.

    “What are you doing?”, asked Minsky.

    “I am training a randomly wired neural net to play Tic-Tac-Toe” Sussman replied.

    “Why is the net wired randomly?”, asked Minsky.

    “I do not want it to have any preconceptions of how to play”, Sussman said.

    Minsky then shut his eyes.

    “Why do you close your eyes?”, Sussman asked his teacher.

    “So that the room will be empty

  • there is a digital modern artist named Jason Salavon that has been working on all sorts of algorithm based art, including a machine that pumps out abstract expressionist paintings [salavon.com]

  • quick! Two hours to go! http://otonx.com/ [otonx.com]
  • Hello!!! everybody, Fashion,low price,the good shoping place, click in. ===== ( http://www.sport3trade.net/ [sport3trade.net] ) ===== Discount Air Jordan (1-24) shoes $35, Air max shoes (TN LTD BW 90 180) $36, Nike/shox (R4, NZ, OZ, TL1, TL2, TL3) $35, Handbags ( Coach Lv fendi D&G) $36, T-shirts (polo, ed hardy, lacoste) $20, Jean (True Religion, ed hardy, coogi)$35, Sunglasses ( Oakey, coach, Gucci, Armaini)$16, Watches(Rolex BREITLING IWC) New era cap $12, Discount (NFL MLB NBA NHL) jerseys, free shipping, Accept cred

In order to dial out, it is necessary to broaden one's dimension.

Working...