![[FREE GAME] Get Distraint Deluxe Edition free](https://i2.wp.com/bestinslot.co/wp-content/uploads/2018/11/distraint-deluxe-edition-free.jpg?resize=360%2C180&ssl=1)
Elon Musk is once again in the news for taking down a team full of DotA 2 pros. Last year his bot defeated the scene’s best player, Dendi in a 1v1 match. Yesterday, on August 6 in San Francisco, a group of DotA 2 pros was taken down by OpenAi bots.
A 5v5 match against bots
Known as OpenAI Five, the bot team collectively took down the human team in a best-of-three series. The human team consisted of former pro players Ben “Merlini” Wu, William “Blitz” Lee, and Ioannis “Fogged” Loucas. The last two players were current pro player, David “Moonmeander” Tan and shoutcaster Austin “Capitalist” Walsh. It’s important to point out that the games weren’t a typical DotA 2 match. There were certain set conditions to cater to the bots current parameters. Only 18 heroes were playable out of the original 115 plus couriers were invincible.
The OpenAI Five won their first and second match easily. During the first match, the human team didn’t even get a chance to take down a tower but they managed to snag one in match two. The human team finally won their last match but this could have been due to the fact that the spectators were the ones determining the bots line up. More than 100k viewers tuned in during this matchup that was live on Twitch.
Bots and the future
As mentioned above, this wasn’t the OpenAI bots first rodeo. Despite the match being under limited conditions, the bots did manage to showcase its capability when it took down Dendi last year. The bots learn how to play DotA 2 via repeated game matches at the rate of 180 years per day in its self-learning simulation. It’s no surprise that it has more than a few tricks up its sleeve. Currently, the bot is set to compete with DotA 2’s best team in this year The International 8. The annual event will begin later this month on August 20. If you’re wondering why these bots are learning to play video games; according to the team:
Systems which solve complex video games will become highly general, with applications outside of games.
With bot-tech advancing at this pace, sooner or later we’ll probably start watching professional bot teams battling out in the arenas.
