OpenAI Five
At OpenAI, we’ve used the multiplayer video game Dota 2 as a research platform for general-purpose AI systems. Our Dota 2 AI, called OpenAI Five, learned by playing over 10,000 years of games against itself. It demonstrated the ability to achieve expert-level performance, learn human–AI cooperation, and operate at internet scale.
Project Timeline
November 9, 2016
First commit of OpenAI's Dota 2 project.
March 9, 2017
First commit in Rapid repository.
August 11, 2017
1v1 Shadow Fiend bot beats top professional Dota 2 players at The International 7.
September 7, 2017
First game won by a Dota 2 professional through normal gameplay against the final 1v1 Shadow Fiend bot after being attempted by dozens of pros over thousands of games.
February 28, 2018
First 5v5 results: RL agent beats OpenAI scripted bot at tower minigame.
April 3, 2018
RL agent beats in-house OpenAI team at net worth minigame.
June 6, 2018
RL agent defeats in-house OpenAI team at fairly restricted 5v5.
#AI bots just beat humans at the video game Dota 2. That’s a big deal, because their victory required teamwork and collaboration – a huge milestone in advancing artificial intelligence.
June 30, 2018
OpenAI Five’s parameters initialized.
August 5, 2018
OpenAI Five defeats popular casters at the Benchmark in front of a live audience and 100k livestream viewers, with somewhat restricted 5v5.
August 9, 2018
Match vs. Team Secret (1–2): OpenAI Five loses its first match to professional players.
August 17, 2018
First significant surgery: OpenAI Five is migrated to Dota 7.19 after training for 7 weeks on Dota 7.16.
August 22–24, 2018
OpenAI Five loses to paiN Gaming and a team of top professional Dota 2 players at The International 8.
August 26, 2018
More model capacity: OpenAI Five’s long short-term memory (LSTM) size is doubled and number of parameters quadrupled.
October 5, 2018
Match vs. Team Lithium (2–0).
December 10, 2018
Extensive surgery to accommodate Valve's 7.20 gameplay update, including map changes, hero changes, and item changes.
January 16, 2019
Match vs. SG e-sports (2–0).
February 1, 2019
Match vs. Alliance (2–0).
April 5, 2019
Final surgery to upgrade to the 7.21d patch, which would not have been possible from scratch.
April 13, 2019
OpenAI Five wins back-to-back games versus Dota 2 world champions OG at Finals, becoming the first AI to beat the world champions in an esports game.
You play against [OpenAI Five] and you realize it has a playstyle that is different. It’s doing things that you’ve never done and you’ve never seen. Sometimes it looks extremely silly. But then again, are you going to be human and be like "Hey, this looks very stupid, this is bad" or [do] you try to take it to next steps, like "Why is it doing this?"
One key learning that we took is how it was allocating resources. It’s just allocating resources as efficiently as possible. And then you realize that we’re guilty of being stuck in a team dynamic, whereas sometimes we have to be way more flexible. […] If OpenAI does that dynamic switch at 100%, we maybe went from 5% to 10%? But that is already a difference—we’ve noticed it.
April 18–21, 2019
OpenAI Five is scaled up to play the Internet as competitor or teammate in OpenAI Arena.
July 12, 2019
Finished training of a new agent, Rerun, which reached a 98+% win rate against the agent that played at Finals. This was completed in 2 months, without surgery, and while utilizing only 20% of the resources.
August 25, 2019
OG wins The International 9, making history as the first two-time world champions.
I don't believe in comparing OpenAI Five to human performance, since it's like comparing the strength we have to hydraulics. Instead of looking at how inhuman and absurd its reaction time is, or how it will never get tired or make the mistakes you'll make as a human, we looked at the patterns it showed moving around the map and allocating resources.
In terms of what OpenAI has done for us and how it influenced our run at TI9, one of the many curious patterns was the buyback and pressure play that happened in most of the games. We had a lot of talks about fighting and pressuring and how it used a different approach from any human in the past. As people, it's about being realistic and learning from the brain of the AI and not the hydraulic strength that machines have.
Selected Press
Team
Christopher Berner
Greg Brockman
Brooke Chan
Vicki Cheung
Przemysław Dębiak
Christy Dennison
David Farhi
Quirin Fischer
Scott Gray
Shariq Hashme
Christopher Hesse
Rafal Józefowicz
Catherine Olsson
Jakub Pachocki
Michael Petrov
Henrique Pondé de Oliveira Pinto
Jonathan Raiman
Tim Salimans
Jeremy Schlatter
Jonas Schneider
Szymon Sidor
Ilya Sutskever
Jie Tang
Filip Wolski
