# Elsewhere — recommended reading

> Other people's writing that Scott Taylor keeps going back to.

Canonical: https://st.im/elsewhere/

## Essays read on a monthly basis

- [The Bitter Lesson](http://www.incompleteideas.net/IncIdeas/BitterLesson.html) — Rich Sutton, 2019: Two pages that keep being right. I reread it whenever a clever hand-built approach starts feeling like the answer, which is roughly monthly.
- [You and Your Research](https://www.cs.virginia.edu/~robins/YouAndYourResearch.html) — Richard Hamming, 1986: On why capable people mostly do unimportant work, and what the ones who don't have in common. Uncomfortable in the useful direction.
- [Maker's Schedule, Manager's Schedule](https://paulgraham.com/makersschedule.html) — Paul Graham, 2009: The clearest account of why a single meeting can cost an afternoon. I reread it whenever the calendar starts winning.

## Books

- Thinking in Bets — Annie Duke, 2018 (currently reading): Separating the quality of a decision from the quality of its outcome. The habit the quant book taught me, written down properly.
- Seeing Like a State — James C. Scott, 1998: How systems built for legibility quietly destroy the local knowledge that made the thing work. Applies to governments, and to roadmaps.

## Essays

- [The Rise of Worse is Better](https://www.dreamsongs.com/RiseOfWorseIsBetter.html) — Richard Gabriel, 1991: The argument behind shipping the boring version, made thirty years before I worked it out for myself.

## Blogs

- [Gwern.net](https://gwern.net) — Gwern Branwen: Essays that are never finished, only revised — each one dated, versioned and argued with itself in public. The closest thing the web has to a working notebook kept honestly.
- [Dan Luu](https://danluu.com) — Dan Luu: Measures the things everyone else asserts. Reliably ends up somewhere less comfortable and more useful than the consensus.
- [Simon Willison's Weblog](https://simonwillison.net) — Simon Willison (currently reading): The most consistent primary source on what language models can actually do this week, written by someone who tries it rather than speculating about it.