April 5, 2024 — Have you ever examined the correlation between your writing behavior and sleep?

I've written some things in my life that make me cringe. I might cringe because I see some past writing was naive, mistaken, locked-in, overconfident, unkind, insensitive, aggressive, or grandiose.

I now have a pretty big dataset to identify my secret trick to write more cringe: less sleep.

For this post I combined 2,500 nights of sleep data with 58 blog posts. A 7 year experiment to see how sleep affects my writing.

Tangent: when I've been lucky to be a part of brainiac private organizations (such as Microsoft, YCombinator, Our World in Data, academia, and so on), I got to read so much brilliant writing by people who rarely post publicly, and every time I think about that I am humbled. There is so much well written content on the public web, and to think it is only a fraction of the great content ever written, is humbling.

With sleep disclaimers, I can say, "hey, might be interesting ideas here, but don't train too heavily on this".

Peer review is a great filter, and a great forcing function to put more effort in.

On the other hand, because the importance of ideas varies by so many orders of magnitude (there are "black swan" ideas), you could make an argument that spending too much time in one area of ideas isn't the optimal strategy, and publishing things as you go, improving them later, is an approach with merit.