3 ms·
(I work at OpenAI.) We'll have another blog post coming in the next few days. But as a sneak peek: we use self-play to learn everything that depends on an inte
by gdb 9y ago
(I work at OpenAI.)
We'll have another blog post coming in the next few days. But as a sneak peek: we use self-play to learn everything that depends on an interaction with the opponent. Didn't need to with those that don't (e.g. fixed item builds, separately learned creep block).
- RSchaeffer 9y agoCan you give us a sneak peak of OpenAI's approach to 5v5 Dota? :D
- hayd 9y agoSo it's not exclusively unsupervised self-play? This contradicts how it was portrayed during the game/afterwards.
- gdb 9y agoSee the opening video (https://www.youtube.com/watch?v=wiOopO9jTZw&t=268 https://www.youtube.com/watch?v=wiOopO9jTZw&t=268), where I mention that we "coach" it on what we think is good and bad. All the heavy lifting is self-play (which I would characterize as supervised).
- nbkvjones 9y agoThat's good to hear. There's probably a huge disconnect between what an AI engineer entails with a certain word and what the layman then comprehends. At the same time, I think we (nadota) probably latched onto certain words and downplayed the ones that would make it more impressive to us. Just a result of itching for information without there being much out there. Really cool work, really glad you're doing it in dota, and eagerly awaiting your blog post.
- Flair 9y agohttps://blog.openai.com/more-on-dota-2 https://blog.openai.com/more-on-dota-2 A new blog is already here.