My Twitter algorithms seem optimitized: a @ghadfield point, @garymarcus contra point, lining up succinctly in my twitterfeed.
A healthy balance of realist skepticism.pic.twitter.com/6UvJO4CQkx
-
Yes it’s the learning of position in space and how to manipulate. Not the puzzle solving
Certainly agree that is where value of the work is. I wish that (a) they had not framed the piece as puzzle-solving, & (b) demonstrated manipulation with multiple objects, without bluetooth sensors inside. Less focused on Rubik's would have been more impressive, less misleading.
-
-
The central challenge between today and when we reach general AI capability will be navigating the public discourse on
#AIAlignment, and dev. of ethical standards. That explainability gap is a real challenge. The public cares about performance, not what's under the hood. -
Yes, AI alignment is core challenge. I think the challenge goes well beyond ethics. And explainability—in sense of knowing what’s under the hood—indeed won’t be sufficient or even in some cases necessary. Key thing is justifiability of AI behavior
End of conversation
New conversation -
-
-
Agreed--I guess I've been impressed by watching the hand in operation with objects and under perturbations such as being poked by a plush giraffe....https://openai.com/blog/solving-rubiks-cube/#perturbations …. To be fair, the
@openai post opens with Rubik's cube but focuses on dexterity -
the giraffe was my favorite part, too
End of conversation
New conversation -
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.