Hacker News Reader

HNHacker News

Top New Best Ask Show Jobs

Comment by kuu | Hacker News Reader

Parent Full thread

kuu·

While I agree with your comment, this sentence:

"This was given to the AI by having it watch a video that explains it."

This was not as trivial as it may seem just a few months ago...

View on HN

EDIT: Incorrect, see below

it didn't watch 'a video', it watched many, many hours of video of playing minecraft (with another specialised model feeding in predictions of keyboard and mouse inputs from the video). It's still a neat trick, but it's far from the implied one-shot learning.

The author replied in this thread and says the opposite.

Ah, I was incorrect. I got that impression from one of the papers linked at the end of the article, but I suspect that's actually some previous work.

I applaud you for acknowledging your mistake. So many people double down, especially in this pernicious and polarized age.

Alpha Star was also trained initially from youtube videos of pros playing Starcraft. I would argue that it was pretty trivial a few years ago.

Do you know if it was actual videos or some simpler inputs like game state and user inputs? I’d be impressed if it was the former at that time.

starcraft provides replay files that start with the initial game state and then every action in the game. Not user inputs, but the actions bound to them.

I don't think it was videos. Almost certainly it was replay files with a bunch of work to transform them into something that could be compared to the model's outputs. (Alphastar never 'sees' the game's interface, only a transformed version of information available via an API)

This was my understanding as well, as the replay files are all available anyway.

The YouTube documentary is actually very detailed about how they implemented everything.

Which documentary? Is it this one?

https://www.youtube.com/watch?v=UuhECwm31dM

It was a ~1h documentary

Reply on news.ycombinator.com