NOTE: Part 2 of this article can now be read here.
Ali Minai at Bounded Alignment:
In the last couple of weeks, there has been a spate of elite voices from deep within the AI enterprise prophesying doom (here’s a good compilation by Taylor Popielarz of Spectrum News). There are bills in Congress to put a moratorium on large data centers, ban the development of superintelligence, and to add a “kill switch” to AI. It seems that we are in the process of crossing a threshold where the idea of AI as an existential threat is breaking through to the public.
Why is this happening now? Why are so many people from within the system speaking out suddenly? To those who pay closer attention, this is not surprising. Predictions of AI doom are, of course, standard sci-fi stuff, but there has been a decades-long tradition of semi-philosophical writing on the topic, captured well by the title of a book by two leading proponents of AI doom: “If Anyone Builds It, Everyone Dies”. Well-known prognosticators have been publishing dire predictive scenarios, and putting out proposals to postpone them by at least a few years. Major AI figures such as Geoff Hinton and Yoshua Bengio have been making frightening predictions for several years. But something has changed very recently.
More here.
Enjoying the content on 3QD? Help keep us going by donating now.