Artificially intelligent game bots pass the Turing test on Turing's centenary. - Polidicks

[quote] [B]An artificially intelligent virtual gamer created by computer scientists at The University of Texas at Austin has won the BotPrize by convincing a panel of judges that it was more human-like than half the humans it competed against.[/B] [quote] ================= [IMG]http://cdn.physorg.com/newman/gfx/news/2012/artificially.jpg[/IMG] ================= UT^2 game bot faces off against an opponent. Credit: Jacob Schrum ================= [/quote] The competition was sponsored by[B] 2K Games [/B]and was set inside the virtual world of "[B]Unreal Tournament 2004[/B]," a first-person shooter video game. The winners were announced this month at the IEEE Conference on Computational Intelligence and Games. "The idea is to evaluate how we can make game bots, which are nonplayer characters (NPCs) controlled by AI algorithms, appear as human as possible," said Risto Miikkulainen, professor of computer science in the College of Natural Sciences. Miikkulainen created the bot, called the UT^2 game bot, with doctoral students Jacob Schrum and Igor Karpov. The bots face off in a tournament against one another and about an equal number of humans, with each player trying to score points by eliminating its opponents. Each player also has a "judging gun" in addition to its usual complement of weapons. That gun is used to tag opponents as human or bot. [B]SEE SOURCE FOR VIDEO[/B] The bot that is scored as most human-like by the human judges is named the winner. UT^2, which won a warm-up competition last month, shared the honors with MirrorBot, which was programmed by Romanian computer scientist Mihai Polceanu. The winning bots both achieved a humanness rating of 52 percent. Human players received an average humanness rating of only 40 percent. The two winning teams will split the $7,000 first prize. The victory comes 100 years after the birth of mathematician and computer scientist Alan Turing, whose "Turing test" stands as one of the foundational definitions of what constitutes true machine intelligence. Turing argued that we will never be able to see inside a machine's hypothetical consciousness, so the best measure of machine sentience is whether it can fool us into believing it is human. "When this 'Turing test for game bots' competition was started, the goal was 50 percent humanness," said Miikkulainen. "It took us five years to get there, but that level was finally reached last week, and it's not a fluke." [B]SEE SOURCE FOR VIDEO [/B] The complex gameplay and 3-D environments of "Unreal Tournament 2004" require that bots mimic humans in a number of ways, including moving around in 3-D space, engaging in chaotic combat against multiple opponents and reasoning about the best strategy at any given point in the game. Even displays of distinctively human irrational behavior can, in some cases, be emulated. "People tend to tenaciously pursue specific opponents without regard for optimality," said Schrum. "When humans have a grudge, they'll chase after an enemy even when it's not in their interests. We can mimic that behavior." In order to most convincingly mimic as much of the range of human behavior as possible, the team takes a two-pronged approach. Some behavior is modeled directly on previously observed human behavior, while the central battle behaviors are developed through a process called neuroevolution, which runs artificially intelligent neural networks through a survival-of-the-fittest gauntlet that is modeled on the biological process of evolution. Networks that thrive in a given environment are kept, and the less fit are thrown away. The holes in the population are filled by copies of the fit ones and by their "offspring," which are created by randomly modifying (mutating) the survivors. The simulation is run for as many generations as are necessary for networks to emerge that have evolved the desired behavior. "In the case of the BotPrize," said Schrum, "a great deal of the challenge is in defining what 'human-like' is, and then setting constraints upon the neural networks so that they evolve toward that behavior. "If we just set the goal as eliminating one's enemies, a bot will evolve toward having perfect aim, which is not very human-like. So we impose constraints on the bot's aim, such that rapid movements and long distances decrease accuracy. By evolving for good performance under such behavioral constraints, the bot's skill is optimized within human limitations, resulting in behavior that is good but still human-like." Miikkulainen said that methods developed for the BotPrize competition should eventually be useful not just in developing games that are more entertaining, but also in creating virtual training environments that are more realistic, and even in building robots that interact with humans in more pleasant and effective ways. =============================== Source: [URL]http://phys.org/news/2012-09-artificially-intelligent-game-bots-turing.html[/URL] [/quote] Pretty impressive. :dance:

[quote] "People tend to tenaciously pursue specific opponents without regard for optimality," said Schrum. "When humans have a grudge, they'll chase after an enemy even when it's not in their interests. We can mimic that behavior." In order to most convincingly mimic as much of the range of human behavior as possible, the team takes a two-pronged approach. Some behavior is modeled directly on previously observed human behavior, while the central battle behaviors are developed through a process called neuroevolution, which runs artificially intelligent neural networks through a survival-of-the-fittest gauntlet that is modeled on the biological process of evolution. Networks that thrive in a given environment are kept, and the less fit are thrown away. The holes in the population are filled by copies of the fit ones and by their "offspring," which are created by randomly modifying (mutating) the survivors. The simulation is run for as many generations as are necessary for networks to emerge that have evolved the desired behavior. [/quote] So they've managed to create self evolving humanlike AI's capable of holding [I]grudges[/I]. thanks guys you're real team players

It'll be amazing when single player games start to pass a Turing benchmark. I imagine the test would be some form of counter-op multiplayer campaign where players are inserted as random AI characters.

If I had the judging gun in HL2:DM, everyone would be marked a bot. Seriously, how do they pull off those gravity-nade headshots in midair with a ping of 999?

maybe they should get people that can actually play the game well to be the judges next time because those videos arent convincing

[QUOTE=longears34;37817387]maybe they should get people that can actually play the game well to be the judges next time because those videos arent convincing[/QUOTE] I doubt that those videos are records of actual testing

I hope that sometime soon this will aid in creating better NPC AI. Being a singleplayer exclusive gamer, I can really appreciate good AI.

[quote]The winning bots both achieved a humanness rating of 52 percent. Human players received an average humanness rating of only 40 percent. [/quote] so the bots are more human than the humans that's a twist i wasn't expecting

I wonder if they also emulate typical fps gamer behaviour "OMFG HOW DID YOU KILL ME YOU HIT NOWHERE NEAR ME U FUCKING HACKER"

They should have a voice synthesis too, so they could call you names. Then it'd be perfect.

Pit it against Brazilian players.

[QUOTE=thisispain;37817459]so the bots are more human than the humans that's a twist i wasn't expecting[/QUOTE] More human than human is our motto.

[video=youtube;RbFL0j-IqLo]http://www.youtube.com/watch?v=RbFL0j-IqLo[/video]

[QUOTE=Warship;37817689]They should have a voice synthesis too, so they could call you names. Then it'd be perfect.[/QUOTE] Microsoft Sam calling you a cocksucker would be hilarious.

So, where can you download these bots? I always found the UT2004 AI to be lacking.

I'd love this kind of bots for Red Orchestra

[QUOTE=DinoJesus;37817964]Microsoft Sam calling you a cocksucker would be hilarious.[/QUOTE] We could play against John Madden! AEIOU SNAKE? SNAKEEEEEEEE!

[QUOTE=Warship;37817689]They should have a voice synthesis too, so they could call you names. Then it'd be perfect.[/QUOTE] I played against a bot on SupCom once that would taunt you and make fun of you as it fought you. "Oh thats a cute sc- well it was a cute scout" "Nice arty base there, shame if something were to happen to it" "I see you...."

[QUOTE=thisispain;37817459]so the bots are more human than the humans that's a twist i wasn't expecting[/QUOTE] h+ Also wow UT2004 wasn't exactly the kind of thing that I expected to be the first thing that passes the Turing test.

[QUOTE=thisispain;37817459]so the bots are more human than the humans that's a twist i wasn't expecting[/QUOTE] We need to make them less human to make them more human.

So they've finally made a bot that isn't a shitbrain or an aimbot on PCP

Always wanted a self evolving AI in a game.

I want a multiplayer shooter that watches how [I]everyone[/I] plays, so that the bots eventually learn by example and become godlike bunnyhopping deathmachines.

Um did anyone watch those videos? I too think these players and bots are very human so human in fact that they play like my grandmother would

[QUOTE=POLOPOZOZO;37818766]Um did anyone watch those videos? I too think these players and bots are very human so human in fact that they play like my grandmother would[/QUOTE] Both of those videos are from the perspective of human judges. And they're terrible.

[QUOTE=BlkDucky;37818971]Both of those videos are from the perspective of human judges. And they're terrible.[/QUOTE] Do these people even play video games? His response to being shot at was to stand completely still out in the open and look around. No wonder these people were mistaken for bots.

[QUOTE=BlkDucky;37818971]Both of those videos are from the perspective of human judges. And they're terrible.[/QUOTE] If that's the case, I wonder if it's a legit Turing test. The original Turing test is about having blind conversations with computers and humans, which can easily be judged by anyone. However a game bot might not be so easily identified by inexperienced players. You need to have a feel for the game to know how people act in it.

[QUOTE=proch;37818083]I'd love this kind of bots for Red Orchestra[/QUOTE] RO is a lot harder... you'd have to make bots that take cover and stuff.

[QUOTE=Clavus;37819135]If that's the case, I wonder if it's a legit Turing test. The original Turing test is about having blind conversations with computers and humans, which can easily be judged by anyone. However a game bot might not be so easily identified by inexperienced players. You need to have a feel for the game to know how people act in it.[/QUOTE] thats what i thought, the test was made by giving the humans players a weapon to tag when they think the target is a human or a bot. looking on the videos the humans players are quite bad on the game. the test would be better if it was using experienced players