David Wallace-Wells
Calls to mind @tobyordoxford’s “scaling paradox” for AI, that even astonishing progress in model capability depends on exponential growth in cost which makes the arc of improvement look (at least) a lot less impressive: https://www.tobyord.com/writing/the-scaling-paradox
Ramez Naam
🧵 If you want to see the diminishing returns in AI, consider that in OpenAI's recent RSI blog post: 124x increase in token output Led to 7x increase in lines of code Led to 1.6x increase in experiments run Led to (my estimate, based on power laws) Perhaps 10% increase in AI capability improvement rate.
