As progress in Deep Learning is made with more complex problems like Dota collaborative play, it will be increasingly difficult to evaluate what the system is actually doing and how it is doing it. Throwing shade will remain a reasonable argument for quite a while.
Since I have thrown some shade on notion of OpenAI’s DoTA system as being strictly comparable to humans, let me also be clear that the work is still interesting; e.g. I agree w @andrey_kurenkov on value in work on complex coordination & long horizon in incompletely observed gameshttps://twitter.com/andrey_kurenkov/status/1031252792487231488 …
-
-
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Did anyone notice that LSTM networks are the choice of algorithm behind the most interesting AI developments ? Without sequences, it would be impossible to make a plan, strategize or solve a problem.
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.