Dwarkesh Patel · 实时热榜

A debate about recursive self-improvement.

Locking in AI safety regulation now is a mistake.

The end of cheap compute?

If a human-level software engineer that could run on an H100 equivalent, at current market rates for software engineers, that H100 should rent for over $250k a year. That’s 15x today’s spot price.

And why black holes are the ultimate power plants.

Abolishing pandemics/ Getting out of the way of AI automation/ Learning from Honk Kong MTR's business model

Watch now (94 mins) | Math is where we’ll see superintelligence first. What will it look like?

Labs are throwing away the most valuable data.

"We see these AIs as a galaxy glittering with capabilities, but at their center, invisible to the naked eye, holding all the constellations together, is an unimaginably massive black hole of data."

"He begged to work for the regime that tortured him."

“One robot now turns into many robots next year, but the number of ballerinas is the same.”

Working up from basic logic gates to why GPUs, TPUs, FPGAs, and the human brain each look the way they do.

If your definition of intelligence is "the ability to achieve your goals across a wide variety of domains", then Stalin was the most intelligent person who ever lived.

Deeply researched interviews

the verification loop for theories can be on the order of decades and centuries, and even then we know today as the better theory can often actually make worse predictions

AlphaGo is still the cleanest worked example of the primitives of intelligence: search, learning from experience, and self-play.

"Instead of being quiescent, natural selection is everywhere."

It's shocking how much you can deduce about what the labs are doing from a handful of equations and a blackboard

Hodge podge of things I was thinking about this weekend.

The not-so-secret point of this whole contest is so that I can hire a researcher