Published: Fri, August 10, 2018
Tech | By Dwayne Harmon

Elite Dota 2 Team Easily Beat by Advanced AI Bots

Elite Dota 2 Team Easily Beat by Advanced AI Bots

It had taken just two weeks for that bot to go from knowing nothing about Dota 2, the battle arena video game from Valve in which the goal is to destroy the oppostion's large structure known as the Ancient while defending their own.

In the three-game series, OpenAI Five bots won the first two games and in the final game, lost it to the human players, as the audience selected their team of five against the AI team.

The human vs machine challenge began with a warmup match against the audience, which OpenAI won easily.

Prior to the Benchmark Game a team of volunteers for the audience played a public match against OpenAI Five.

Sunday's tournament was a test run before the world Dota 2 championships later this month, where the bots will compete against a team of professional players.

And the OpenAI has the advantage of learning through rigorous reinforcement training, a machine learning technique that involves teaching AI agents through trial and error and then "rewarding" them when they carry out a good action in the game.

Dota 2 is available now for PC, Mac, and Linux. It predicted a 95 percent win probability after seeing the hero teams. Brockman added that to overcome these challenges, the bots had to be able to develop "intuitions" about their human opponents, improvise in response to unfamiliar situations, and collaborate with one another. In preparation the bots played through 180 years' worth of games against each other every day. The firm has also been tweaking the neural networks, altering the bot team's in-game reaction time to near human levels, and helping it learn additional strategies. The team of bots calculated a 2.9% chance of winning before the start of the game based on its heroes' lineup.

OpenAI Five's latest win against its most formidable adversaries yet ultimately demonstrates the enormous progress the bots have made over the past few months.

"It's hard to know what kind of progress you're making if you're just making progress in simulators", Jack Clark, who works at OpenAI, said in an interview with Axios.

Like this: