PPO (Proximal Policy Optimization)
A reinforcement learning algorithm that trains agents by having them plan over every single time step without hierarchy; used to train DOTA systems and described as flawed but scalable.
As defined in OpenAI Co-Founder: “I just had to Quit”
Where it comes up · 1
Most important claims · 1 of 1Search all claims for “PPO (Proximal Policy Optimization)” →
- DOTA was a watershed moment not because it was chess-like (Deep Blue) or computationally intensive with defined rules (AlphaGo), but because it involved human-like interactive agency in a partially-defined, continuous environment—demonstrating that massive compute + simple algorithms could exceed human-level performance in real-world complexity.OpenAI Co-Founder: “I just had to Quit”