0: I got out of the habit of writing these posts because I’m trying to be on my phone less. But I want to get back into it. It’s still interesting to review these. “The world” is moving fast this year.
1: Dean Ball on the Ezra Klein show — likely evergreen. By the way there will always be jobs if there is a human that can be sued for every agent action.
2: Emmanuel Saez and Josh Rauh debate — on the permanently experimental California wealth tax.
3: Shyam Sankar on Invest Like The Best — the maximal rate of learning is coincident with how much pain you are handling. Comforting career ladders don’t maximize learning.
4: Jack Clark on Plain English & Terence Tao on The Dwarkesh Podcast — is “ambient creativity” unique to humans? Researchers at the frontier consistently say you need time at rest to be creative. “You waste years by not being able to waste hours” etc. Creativity is another reason there may always be jobs.
Using a strict keyword classification methodology applied to a 100,000-document sample of The Pile — a foundational open-source pretraining corpus of approximately 825 billion tokens — we estimate that approximately 67 billion tokens (8.1% of the corpus) consist of explicitly Christian content: scripture, theology, sermons, catechisms, apologetics, and church history. This figure exceeds Islamic content by 32x, Buddhist content by 19x, Hindu content by 45x, and Jewish content by 93x. It is equivalent to 15 times the entire English Wikipedia, and exceeds the estimated total output of the AI safety research field by two to three orders of magnitude.”
6: Quantization from the ground up — There is a lot to learn.
7: A Directory of Living Literature Reviews — Synthesis is relatively more important as collective knowledge increases.
8: What will be scarce? — Great article.
9: Will AI kill the research paper? — also a banger. I almost applied to econ grad school. I still go back and forth between very glad I didn’t, and wondering if I should have. Maybe I should just start working on some econ “boxes”.
10: Tyler Cowen at ARC 2026 — “One thing that is not true: AI will not bring mass unemployment.”
11: Lore — Individualized cycling shoes. I want to make this but for durable running shoes.
12: the old world is dying by Jasmine Sun
First, no nostalgia! The old world was bad. Those staid institutions and rigid hierarchies are not worth mourning. I went to a grinder high school and a grifter college. Both were mostly shitty and anti-human ways to live. Nothing makes you feel as compressed, quantized, and box-checked as AP season or the college application process. The notion that teens can smoke weed and skim Hegel for four years and have a cushy office job waiting on the other side was always doomed. Did new grad engineers have a God-given right to a $200k starting salary, while schoolteachers and paramedics grind for a fraction of the wage? There have to be better credentialing systems than one’s willingness to go into debt.
13: Speed Can Reindustrialize America — this essay is living rent free in my head.
14: Build, Baby, Build: How Housing Shapes Fertility by Benjamin Couillard — “To study the causal effect of rising housing costs on fertility, I vary them directly within the model, finding that rising costs since 1990 are responsible for 11% fewer children, 51% of the total fertility rate decline between the 2000s and 2010s, and 7 percentage points fewer young families in the 2010s.”
15: Yep.
People who say "we need to figure out a plan to prevent AI from putting people out of work" don't seem to understand that what they're really saying is "we need a plan to ensure AI doesn't drive economic growth." They're two sides of the same coin.
— Timothy B. Lee (@binarybits) April 8, 2026
16: Vital Lyfe — easier than ever to build a modern, off-grid home. And at what point is it easier/cheaper to build your own home given construction costs? Long hobbyists.
17: Dean Ball on the encyclical. I agree.
Humanity is building machines that will be smarter than we are at things we care about, things in which take individual and collective pride, domains of thought we originally invented and discovered. This will enable incredible things, but no honest person can deny that this will…
— Dean W. Ball (@deanwball) May 25, 2026
18: Europe 2031 — obligatory mention. I would like to see Europe build again!
Disclaimer: Something something sharing a link is a recommendation but not necessarily an endorsement or a sign of agreement something something.