Deeper analysis of the AI vs Human dogfight. Showing that AI systems will always find a way to hack the normal/accepted human rules of any game. They win not because they think like us - they win because they are fundamentally a different kind of intelligence. https://twitter.com/ZachFB/status/1296838707849375745 …
-
This Tweet is unavailable.
-
Replying to @sgourley
Some of the best advice I ever got in ML courses: when an advance in sim envs come this readily, it’s because your reward function isn’t tracking anything real.
1 reply 1 retweet 1 like
Replying to @WillManidis @sgourley
and in this case, there’s a huge difference between the reward function of human: win, don’t die, don’t lose the plane rl-agent: win
0 replies
1 retweet
1 like
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.