Ryan Greenblatt's case is conditional: AI systems that match leading researchers on verifiable AI R&D tasks could help build stronger successors. The disputed point is whether such gains transfer from bounded training environments to the load-bearing parts of real research—and how reward hacking must be governed along the way.
Adam Brown opens with a problem of explanation. In the lecture's telling, Newtonian gravity looks instantaneous, while special relativity treats the speed of light as a limit on influence.
A high-profile result in machine problem solving would still be one benchmark among others, not the moment general intelligence is settled. Grant Sanderson uses olympiad-style tasks to show capability spreading unevenly across problem types, and asks what separates a correct answer from mathematical understanding.
Ada Palmer opens with the world of The Prince: a divided Italy, unstable Florentine institutions, and a writer addressing Lorenzo de' Medici with a final plea to lift Italy out of ruin. Read back in that crisis, she argues, the book reads less like a manual for tyrants than like a response to catastrophe.
Sarah Paine's framework starts from geography and security conditions, not from a law of history. Continental and maritime powers have organized the world differently, she argues, and the distinction still shapes what each kind wants from order, trade, and war—an ideal type for reading Putin's Russia and Xi's China.
