News for hackers. Stuff that matters.
Re:OpenAI Five Benchmark: Results by ufo on Monday August 06, 2018 @07:50PM (#17702776)
According to the Q&A at the end of the event, one of the main obstacles for using the regular game as output instead of the bot API is that self play would become prohibitively expensive. The AI plays thousands and thousands of games every day and you would need an enormous amount of GPU resources to render all of those.
Re:OpenAI Five Benchmark: Results by skgoa on Tuesday August 07, 2018 @03:00AM (#17704377)
That's a bullshit argument. You can seperate the CV portion from the analysis portion and just train the analysis portion by giving it realistcally limited information.
Re:OpenAI Five Benchmark: Results by ambicapter on Monday August 06, 2018 @09:31PM (#17703254)
That's just equal footing, I say.
Re:OpenAI Five Benchmark: Results by ufo on Monday August 06, 2018 @09:37PM (#17703273)
To be pedantic, equal footing here would be about GPU resources needed to interpret the images. The point I mentioned was about the GPU resources needed to render them to the screen in the first place.