about half of the AI experts at our Puerto Rico conference guessed that it would happen by 2055. At a follow-up conference we organized two years later, this had dropped to 2047.
My guess is that we will have something resembling AGI by the late 2030's. What I mean by this is that it probably won't meet all critera for being AGI but it will appear like AGI to many people and perhaps pass new "Turing tests". Put me in the full AGI and then Super AI by 2050 group.
