I recently heard Connor Leahy give what he described as his initial 5-minute pitch for politicians and other powerful people, and it seemed to very accurately laser in on the x-risk-relevant issue: superintelligence is not here yet, we don’t know how to control it, having superintelligence we can’t control would be extremely dangerous, we should not make it without knowing how to control it. Based on that I have some confidence in what they’re currently doing with communication.
maia
Karma: 3,712
This is fair—obviously the presentation he gave to me and a group of x-risk-concerned folks could differ from what ControlAI is saying to politicians in general. Do you have any more recent public info you can point to about this by any chance? The post you linked is from 2023.