Can AI Research Explode Before Robots Do Physical Work?

Leopold Aschenbrenner says an AI research explosion can start before robots do ordinary physical work. AI research does not need anyone to plow a field or dig up coal, so it skips the real-world bottlenecks that slow other jobs. A system that cannot flip a burger can still make algorithmic progress, and algorithmic progress is what later figures out a robot that flips the burger.

He calls 100 million copies a lower bound we can definitely run, and says one of the first projects is turning quantity into quality. Even at the baseline pace, four years separate the preschooler from the high schooler, and smarter systems arrive quickly. Simple algorithmic changes, he says, are likely: one Alec Radford becomes 100, with no need for 100 million copies. The next systems can do creative, complicated work. He adds that test-time compute might be pooled into one process instead of spent only on parallel copies.

The result is not only a quantitative lead over humans. Leopold Aschenbrenner says the systems become qualitatively superhuman, fast. A high school student still trying to grasp standard physics cannot follow a professor for whom quantum physics is obvious. Automated AI research reaches that gap even faster than the underlying pace of progress.