Back
OpenAI Five | OpenAI
webCredibility Rating
4/5
High(4)High quality. Established institution or organization with editorial oversight and accountability.
Rating inherited from publication venue: OpenAI
OpenAI Five is a key capabilities milestone illustrating what large-scale RL can achieve; relevant to AI safety discussions about capability jumps, emergent behavior, and the compute-performance relationship.
Metadata
Importance: 55/100blog postprimary source
Summary
OpenAI Five was a reinforcement learning system that achieved superhuman performance in Dota 2, a complex real-time strategy game, by training using self-play at massive scale. It demonstrated that large-scale RL with sufficient compute could master long-horizon, multi-agent cooperative and competitive tasks previously considered intractable. The project served as a landmark capabilities demonstration and provided insights into emergent teamwork, strategy, and scaling.
Key Points
- •Achieved superhuman Dota 2 performance, defeating world champions OG at OpenAI Five Finals in April 2019.
- •Trained via large-scale self-play reinforcement learning, accumulating roughly 45,000 years of gameplay experience.
- •Demonstrated emergent cooperative behaviors among five AI agents without explicit coordination mechanisms programmed.
- •Highlighted the role of compute scaling in enabling AI to tackle complex, long-horizon multi-agent environments.
- •Provided lessons relevant to AI safety regarding reward shaping, multi-agent dynamics, and policy robustness.
Cited by 1 page
| Page | Type | Quality |
|---|---|---|
| Deep Learning Revolution Era | Historical | 44.0 |
Cached Content Preview
HTTP 200Fetched Apr 9, 202620 KB
OpenAI Five
Apr
MAY
Jun
01
2023
2024
2025
success
fail
About this capture
COLLECTED BY
Collection: The Markup
TIMESTAMPS
The Wayback Machine - http://web.archive.org/web/20240501161434/https://openai.com/research/openai-five
Close
Search Submit
Skip to main content
Site Navigation
Research
Overview
Index
GPT-4
DALL·E 3
Sora
API
Overview
Pricing
Docs
ChatGPT
Overview
Team
Enterprise
Pricing
Try ChatGPT
Safety
Company
About
Blog
Careers
Residency
Charter
Security
Customer stories
Search
Navigation quick links
Log in
Try ChatGPT
Menu
Mobile Navigation
Close
Site Navigation
Research
Overview
Index
GPT-4
DALL·E 3
Sora
API
Overview
Pricing
Docs
ChatGPT
Overview
Team
Enterprise
Pricing
Try ChatGPT
Safety
Company
About
Blog
Careers
Residency
Charter
Security
Customer stories
Quick Links
Log in
Try ChatGPT
Search
Submit
Research
OpenAI Five
Our team of five neural networks, OpenAI Five, has started to defeat amateur human teams at Dota 2.
June 25, 2018
Dota 2, Reinforcement learning, Self-play, Games, Software engineering, OpenAI Five
Our team of five neural networks, OpenAI Five, has started to defeat amateur human teams at Dota 2. While today we play with restrictions, we aim to beat a team of top professionals at The International in August subject only to a limited set of heroes. We may not succeed: Dota 2 is one of the most popular and complex esports games in the world, with creative and motivated professionals who train year-round to earn part of Dota’s annual $40M prize pool (the largest of any esports game).
OpenAI Five plays 180 years worth of games against itself every day, learning via self-play. It trains using a scaled-up version of Proximal Policy Optimization running on 256 GPUs and 128,000 CPU cores—a larger-scale version of the system we built to play the much-simpler solo variant of the game last year. Using a separate LSTM for each hero and no human data, it learns recognizable strategies. This indicates that reinforcement learning can yield long-term planning with large but achievable scale—without fundamental advances, contrary to our own expectations upon starting the project.
To benchmark our progress, we’ll host a match versus top players on August 5th. Follow us on Twitch to view the live broadcast, or request an invite to attend in person!
Play video
The problem
One AI milestone is to exceed human capabilities in a complex video game like StarCraft or Dota. Relative to previous AI milestones like Chess or Go, complex video games start to capture the messiness and continuous nature of the real world. The hope is that systems which solve complex video games will be highly general, with applications outside of games.
Dota 2 is a real-time strategy game played between tw
... (truncated, 20 KB total)Resource ID:
e388403add7d489b | Stable ID: sid_V5q0n9d5A0