4 gravitons

Do good scholars read every paper they cite? Depends on what the citation is for:

If you want to hold scientists to account, you want them to collaborate internationally.

Why are hallucinated citations possible in the first place? Especially on a platform like arXiv?

"Make no mistakes" isn't a totally useless thing to tell an LLM. But it doesn't do what you think it does.

Some bonus info fo my New Scientist piece last week:

The Breakthrough Prize: I recognize a few of these names!

As promised, my "bonus info" post on my jamming article with Quanta Magazine:

Everything that can happen will happen. Thus, we can learn something about everything.

Some more thoughts on "AI Physicists":

Old news, with my thoughts on it:

Scientists trust what they can verify. Journalists have to trust differently.

If you see a physics paper crammed into six pages of two-column text, here's why:

A few people have asked me about this paper. This is a long piece, but probably not all you were looking for.

In which everyone is wrong about teaching:

favorite word?

Could we mechanize scientific creativity? Not with these papers!

A prose take on last week's topic: