The bot didn't have access to anything the humans couldn't see, but admittedly because it didn't have to move a camera around the screen it was more efficient. They addressed this in the blog post with a bot limited to a camera view (last chart here):https://deepmind.com/blog/alphastar-mastering-real-time-strategy-game-starcraft-ii/ …
deepmind’s latest publicity stunt, dissected.https://twitter.com/MaxALittle/status/1090676568115040258 …
-
-
-
it’s worth reading the whole article
@MaxALittle linked to - 5 more replies
New conversation -
-
-
As long as it is a closed system with strict rules, it can be generalized to a mathematical equation -- and is therefore, not AI, but just a complex coefficients calculator.
-
but "being able to overcome raw numerical complexity via trickery and brute forcing" just doesn't sound quite as sexy as "next generation of AI" now, does it?
End of conversation
New conversation -
-
-
sounds like DRL would benefit from integration with a map-building algorithm.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
"The agent demonstrated strategies I hadn’t thought of before, which means there may still be new ways of playing the game that we haven’t fully explored yet” - the fruits of brute force search.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.