The computers are coming. They’re coming for your jobs. They’re coming to be your lover. Now they’re coming to destroy you in video games (Hopefully they don’t evolve and learn to talk shit about your sweet mother). An Elon Musk-backed research lab named OpenAI built a bot that could play the hugely successful video game Dota 2. Not only could the robot play Dota 2, but it can wreck a professional gamer.
The bot from OpenAI, the $1 billion artificial-intelligence research non-profit that is co-chaired by Musk, took on professional gamer Danylo “Dendi” Ishutin in a one-on-one exhibition matchup at Valve’s flagship $24 million tournament. The first game of the man vs. machine matchup ended after about 10 minutes, then Dendi resigned from the second match after a few minutes, and got his ass whooped so severely that he forfeited to play a third match to prevent any more disappointment. “This guy is scary,” Dendi repeated during play as the computer annihilated him with ease. This proved that the incredibly adaptive AI bot could beat a professional e-sports player in a complex game.
OpenAI first ever to defeat world's best players in competitive eSports. Vastly more complex than traditional board games like chess & Go.
— Elon Musk (@elonmusk) August 12, 2017
OpenAI explains the bot was trained by playing with itself. No. Not like that. Over two weeks, the OpenAI bot played a version of itself in Dota 2 that had the same abilities and learned from mistakes and developed strategies. “We just let it play lifetimes of 1v1 against itself, and coached it on what we thought was good or bad,” OpenAI CTO Greg Brockman said.
This comes after Musk said AI could destroy the world. “I have exposure to the very cutting edge AI, and I think people should be really concerned about it,” Musk said last month. “I keep sounding the alarm bell, but until people see robots going down the street killing people, they don’t know how to react, because it seems so ethereal. AI is the rare case where I think we need to be proactive in regulation instead of reactive because I think by the time we are reactive in AI regulation, it’ll be too late. AI is a fundamental risk to the existence of human civilization.”
OpenAI’s next step is to train five bots to play the popular multiplayer online battle arena (MOBA) game so the robots could compete in a full 5-vs-5 Dota 2 game in next year’s tournament.
Here is the entire event.
[TheVerge]