Nilay Patel

Then again, maybe we shouldn't be so surprised given recent performance.

Both OpenAI and DeepMind entered LLMs in the International Math Olympiad this summer and were capable of gold medal performances.

The notable thing is that these kinds of theoretical math problems have basically zero calculation.

They're about manipulation of logic to come up with a mathematical proof.

It's basically an entirely different category of scientific work.

Former quant investor Jeffrey Emanuel highlighted another interesting novel math paper that required a lot of manual labor to come up with a result.

In a long thread, he suggested that this could be an example of a hidden discovery.

A novel result that was already feasible based on current knowledge, but required too much work for a human to reasonably obtain as an individual or an academic team.

Which gets us to another point.

A recent article in Frontiers was called 90% of science is lost.

And the broader point is that while modern science is about people with 20-year academic careers of extreme specialization, often the largest scientific breakthroughs are about combining observations across fields.

A ton of the big discoveries of the 20th century were, for example, chemistry-slash-physics or biology-slash-physics.

As Frontiers puts it, most scientific data never fuel the discoveries they should.

For every 100 datasets created, around 80 remain in the lab.

20 are shared but rarely used, and only one typically drives new findings.

The result?

Delayed cancer treatments, climate models short on evidence, and research that cannot be reproduced.

That is exactly the type of information that AI could be using and potentially putting to better efforts.

Andrew Curran recently had an interesting post on Twitter where he wrote, We're in a strange spot right now with AI.

The anti-AI crowd believes progress has halted and are doing a victory lap.