Hinton came out with a new idea recently. He's been a bit in the doomer camp but is now talking about a mother baby relationship where a super intelligent AI wants to look after us https://www.forbes.com/sites/ronschmelzer/2025/08/12/geoff-h...

I agree though that much of the debate suffers from the same problem as much philosophy that a group of people just talking about stuff doesn't progress much.

Historically much progress has been through experimenting with stuff. I'm encouraged that the LLMs so far seem quite easy going and not wanting to kill everyone.