I think about this everyday. It’s been a month but the scale of preventable suffering here continues to be unfathomable to me.

Look it's @jacobtref.bsky.social dropping three of the best blogposts of 2025. “What does AI progress mean for medical progress?” gives a great overview. Then it turns into a choose-your-own adventure to read about the bottlenecks AI won't solve *and* where it can still help.

Fascinating paper under fairly realistic conditions but the sample is fairly small. But never underestimate the ability of people to convince themselves the way to do something that feels easiest is actually best all things considered.

Ending USAID is a disaster on a scale I can barely comprehend. For context 14 million people is more than the population of most countries today. It would be multiple times larger than the number of people who died in the Holomodor.

For context, this is kid me from the time I got tested for lead. He would want every kid to have a lead free future. (He would also want some cake, what he's celebrating here. But who among us...)

I still remember getting my blood drawn for lead testing as an elementary schooler in Chicago. I was just under the intervention threshold of 10 mcg/dL. Many of my classmates weren't so lucky. Today, lead poisoning continues to afflict 800 million kids worldwide. 🧵

I'm excited to announce I have a Substack: Charity for All. It's a blog about charity, the people in that space, and the broad political and philosophical landscape that shapes how we think about these topics. substack.com/home/post/p-...

I feel like podcast episodes get longer as the show goes on, so someone should start a prediction market on how long the first 5 episodes will be on average. Non-zero chance episode 5 could be like 15 hours (of awesomeness).

Resetting the sign with “days since Scott Alexander did a racism” back to 0.

Dean Karlan is a huge loss for USAID and, as a result, tens of thousands if not millions of people around the world. He’s dedicated his career to evidence-based cost-effective interventions. If this was actually about efficiency he’s exactly the kind of person you would want around.

Nazi salutes are very bad actually. As Ganz notes Hanania’s piece defending them is basically this: ‘“I can tell you all about Nazis, trust me, I used to be one.” Excuse me one moment, please? You said, what now?’

2024 is wrapping up (or at least I'm off from work) and I'm really proud of all the impactful work Rethink Priorities did this year. I thought it'd be good to cover some my favorite pieces we did this year. 1/n

This article highlights a few reasonable concerns with EA alongside many deeply unreasonable ones. I think I'd like to point out a few issues. (1) One big issue while it implicitly complains that EA doesn't take the concerns of recipients seriously, this is both false...