Benjamin Todd argues transformative AI could arrive by 2030, compressing decades of progress into months

We also raised various worries about the "Transformative AI is imminent" picture:

- Impressive benchmarks don't translate into real-world economic usefulness - Why would AI research automate before ordinary white-collar work? - Hardware bottlenecks might cap the feedback loops - Populist and political backlash could block change - Is 80,000 Hours/Effective Altruism placing too much of a reputational bet on imminent transformative AI? - Does the AGI-pilled community show some troubling signs of being an echo chamber?

Video here: https://www.youtube.com/watch?v=mr8C1eGumQM (2/)

@dioscuri @danwilliamsphil No! But also wow this is the fastest I've seen a podcast launched.

@danwilliamsphil @ben_j_todd Great talking to you today @ben_j_todd! Also, completely trivial, but I didn't find a good moment to raise this in the podcast - has anyone told you that you look a bit like Tom Hiddleston?

@dioscuri @ben_j_todd We should have built up to that as the final question