Nothing shocking in here, but a nice anthropological summary.
I also like the concept that there were always ambitious humans, but forager cultures had measures to restrain them, whereas modern culture empowers them.
"although specialized pipelines such as o1-ioi yield solid improvements, the scaled-up, general-purpose o3 model surpasses those results without relying on hand-crafted inference heuristics."
"Overall, these results indicate that scaling general-purpose reinforcement learning, rather than relying on domain-specific techniques, offers a robust path toward state-of-the-art AI in reasoning domains, such as competitive programming."
My default worry is that ML will increase inequality. It basically contributes more valuable forms of capital, and the return on capital long-term does not seem to go down much as its supply increases (according to Piketty, at least).
But this isn't a given at all, and there are tantalizing data points that suggest ML could push in the other direction, as TFA notes.
Overall I agree with this quote from TFA, that it's not a prophecy, but a choice we can make:
>"We have choices about how we design our systems." With the right design of AI systems (and also public policies), Brynjolfsson says, AI might be able to usher in more widely shared prosperity.
also:
>"And one of my pleas to business managers and to technologists is to think harder about using AI to augment humans and to benefit humanity more broadly rather than have all of the benefits accrued to a very small group."
I'm trying to keep on the lookout for such applications, so if you know of good ones, mention 'em!
Let's not forget he's also discussing things on communities like HN, where I calculate 3 comments/day over the last month (based on a calc I just made, since I subscribe to his comments via https://hnrss.github.io/).
reply