Twitter | Pretraživanje | |
OpenAI
We've observed AIs discovering complex tool use while competing in a simple game of hide-and-seek. They develop a series of six distinct strategies and counterstrategies, ultimately using tools in the environment to break our simulated physics:
Reply Retweet Označi sa "sviđa mi se" More
OpenAI 17. ruj
Odgovor korisniku/ci @OpenAI
Unexpected and surprising behaviors included box surfing, where seekers learn to bring a box to a locked ramp in order to jump on top of the box and then “surf” it to the hider’s shelter.
Reply Retweet Označi sa "sviđa mi se"
OpenAI 17. ruj
Odgovor korisniku/ci @OpenAI
And seekers learn that if they run at a wall with a ramp at the right angle, they can launch themselves upward.
Reply Retweet Označi sa "sviđa mi se"
hardmaru 18. ruj
Odgovor korisniku/ci @OpenAI
Congrats on this very inspiring work! I'd be interested to see this line of work scale up in a way proposed in your previous paper on Neural MMO paper:
Reply Retweet Označi sa "sviđa mi se"
Joseph Suarez 18. ruj
Odgovor korisniku/ci @hardmaru @OpenAI
I desire nothing more than the opportunity to do so. Have been hard at work on Neural MMO since the v1.0 release -- lots of progress since:
Reply Retweet Označi sa "sviđa mi se"
H2O Delirious 19. ruj
Odgovor korisniku/ci @OpenAI
Always Red Vs Blue LOL This is actually amazing to see them learn techniques to one up each other.
Reply Retweet Označi sa "sviđa mi se"
Jakob Foerster 17. ruj
Odgovor korisniku/ci @OpenAI
This is great and looks really cool. Any ideas for how to compare this to complex sequences of moves emerging in other self-play settings such as Go or Dota?
Reply Retweet Označi sa "sviđa mi se"
Amardeep Singh Sidhu 17. ruj
Odgovor korisniku/ci @OpenAI
Hahaha Hiders taking away all the blocks and ramp is my favourite part🤣🤣🤣
Reply Retweet Označi sa "sviđa mi se"
Adam 18. ruj
Odgovor korisniku/ci @adssidhu86 @OpenAI
The best was block surfing !
Reply Retweet Označi sa "sviđa mi se"