I think it's weird when someone says that the chance of stopping AI development everywhere in the world is >1%. It's higher EV to try to decrease p(doom) from 25% to 20% by throwing all our best and brightest at the problem rather than praying for a miracle pause.
It's weird when someone says "I think my complicated idea for preventing destruction of the Earth has some chance of working" and doesn't add "but it'd be crazy to gamble civilization on that."
2
177
Eric retweeted
“Our models perform equivalently to backprop on ML benchmarks, while utilising only local and (mostly) Hebbian plasticity. Our method raises the potential that standard ML algorithms could in principle be directly implemented in neural circuitry.” 🧠
openreview.net/forum?id=Pdau…
