flâneur — a map of the web's best reading

Where the Bitter Lesson ends | the singularity is nearer

geohot.github.io · 583 words · saved by 1 readers

Humanity only has one engineering project, building better engineers than humans. After that, the thing we built can do the engineering. Clips have been making the rounds on Twitter from my second Lex about the “bishop guy” in a chess engine, or a “cone guy” in a self driving car. These engineering ideas look ever more ridiculous. Since the beginning of comma, I wanted to make a machine that could drive cars like a human. Obviously there’s no reference to traffic cones inside human DNA, they learn about them from data. So there shouldn’t be any reference to traffic cones in your codebase. Rich Sutton stated this most iconically in 2019. One thing that should be learned from the bitter lesson is the great power of general purpose methods, of methods that continue to scale with increased computation even as the available computation becomes very great. The two methods that seem to scale arbitrarily in this way are search and learning. But then where does it stop? Why draw the line at DNA

Humanity only has one engineering project, building better engineers than humans. After that, the thing we built can do the engineering. Clips have been making the rounds on Twitter from my second Lex about the “bishop guy” in a chess engine, or a “cone guy” in a self driving car. These engineering ideas look ever more ridiculous. Since the beginning of comma, I wanted to make a machine that could drive cars like a human. Obviously there’s no reference to traffic cones inside human DNA, they learn about them from data. So there shouldn’t be any reference to traffic cones in your codebase. Rich

Explore this link on the map →

related reading