Rob Bensinger ⏹️ “It should be obvious to everyone that OpenAI staff and leadership have OpenAI's social impact as their highest priority”

The fact that the people you are talking to not only think that this is true but also think that it should be *obvious* is a bit wild to me.

I suspect…

Do you think you have solved philosophy?

Jesus christ.

Recent research justifies a concern that AI could escape our control and cause human extinction. Very advanced long-term planning agents, if they're ever made, are a particularly concerning kind of future AI. Our paper on what governments should do just came out in Science.🧵

Yann LeCun: 'Good AIs are going to protect us from bad AIs'

The protection on Messenger currently: