In this controversial Web Summit talk, I argue that #AGI is unnecessary, undesirable & preventable - while tool AI can give us basically all of AI's exciting benefits, and the "but China" argument is flawed.

@WebSummit Be aware..

@WebSummit Your position is unreasonable. AGI is not an existential threat to humanity, and no matter how much money you and the rest of the effective altruism cult spend on this, you’re not going to get traction in either national or international law. The game is over.

@WebSummit Well said! 👏🏼👏🏼

@WebSummit Humans are the bottleneck in an automation process. Having an AGI in their place makes it scalable and accelerates progress. If we want to reach the stars and create Dyson Spheres within this century, we need AGI.

@WebSummit No, we don't need it: tool AI can also build Dyson spheres

@WebSummit It's coming regardless though, so we should prepare right?

@WebSummit LOL: I have a whole segment of my talk debunking the myth that it’s inevitable

@WebSummit Good talk, and agreed.

@WebSummit not only is the "but China" argument flawed, casting it as a race with only one winner can become a self-fulfilling prophecy. ppl should look at historical "the missile gap" and "bomber gap" debates to see how faulty arms-race dynamics perpetuate.

@WebSummit Love these examples from the world of biotech, which show how we can and should also guarantee safety in AI development!