🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Really interesting paper from a great team. Consistent with theories suggesting a high barrier to entry in science - even smart, well-published scientists (or 'well-discovered' inventors) do much worse when they leave their core knowledge domain.

Cool paper: GPT-4 increases hiring success for teachers in rural Ghana by ~70%, compared to human reviewers. Having humans work with GPT-4 *doesn’t* improve performance.

I understand the constraints but 1) universities should be directionally more like this, and 2) the fact that this happens w/in 50 sq miles of SF uniquely in the world is a massive advantage. Hardworking hippie hacker weirdos + $, the Bay Area secret since What the Dormouse Said.

Brian is great as always on a straightforward calculation: what fraction of grid can be solar before you run into issues/"it's too expensive b/c of backup you need". Answer: 40-80% depending on batteries, slightly higher in medium run as tech develops to eat extra load usefully.

Greed.

Correct. Especially for services, due to within company transfer pricing related to tax codes, trade deficits are even more meaningless. If you want to discuss tax arbitrage, that would actually be a reasonable complaint for the USG to make, though!

A must read from Erik. Tldr: "Too hard for courts using present law/interpretation to prevent a lot of market power, but whether a given market action is bad cannot be answered with simple rules, so we still need proper economic analysis for antitrust to be useful."

Hayek’s point is about the fallacy that planners can plan scientifically an economy. Its hubris and its wrongheaded. That included planning via tariffs, industrial policy, industrial subsidies, etc. You know, all the stuff Oren Cass wants.