Cody Fenwick

How Many Lives Does US Foreign Aid Save? We present some estimates for lives saved by US assistance worldwide, with illustrative estimates by recipient country. Our core estimates are for deaths prevented from HIV/AIDS, vaccine-preventable i...

Those estimates match up well with the separate, independent estimates of @charlesjkenny.bsky.social and @justsand.bsky.social at @cgdev.org

They estimate closer to three million per year —>

This seems impossible to adhere to in practice. Countless claims of educational value could undermine someone’s religious beliefs.

entitled to a preliminary injunction. A government burdens the religious exercise of parents when it requires them to submit their children to instruction that poses "a very real threat of undermining" the religious beliefs and practices that the parents wish to instill. Wisconsin v. Yoder, 406 U.S. 205, 218 (1972). And a government cannot condition

Wow that’s broad. Incredibly broad.

www.supremecourt.gov/opinions/24p...

Jury Theorems for Peer Review Marcus Arvan, Liam Kofi Bright, and Remco Heesen Abstract: Peer review is often taken to be the main form of quality control on academic research. Usually journals carry this out. However, parts of maths and physics appear to have a parallel, crowd-sourced model of peer review, where articles are posted on the arXiv to be publicly discussed. In this article we argue that crowd-sourced peer review is likely to do better than journal-solicited peer review at sorting articles by quality. Our argument rests on two key claims. First, crowd-sourced peer review will lead on average to more reviewers per article than journal-solicited peer review. Second, due to the wisdom of the crowds, more reviewers will tend to make better judgements than fewer reviewers will. We make the second claim precise by looking at the Condorcet jury theorem as well as two related jury theorems developed specifically to apply to peer review.

Paper is finally up and open access (www.journals.uchicago.edu/doi/10.1086/...), it's a sequel to an earlier paper where we'd argued that there's not good evidence that pre-publication peer review is a net benefit (www.journals.uchicago.edu/doi/10.1093/...). So in this one we suggest an alternative.

There is a Set of Men lately sprung up amongst us, who endeavour to distinguish themselves by ridiculing every Thing, that has hitherto appear'd sacred and venerable in the Eyes of Mankind. Reason, Sobri-ety, Honour, Friendship, Marriage, are the perpetual Subjects of their insipid Raillery.

Hume also encountered a perennial type of guy in the humanities and also found them annoying.

Hilariously, there’s a whole episode where they discover Tuvok created a holodeck program to simulate a Maquis mutiny, demonstrating the dramatic potential, but he explicitly says he stopped developing the program because it was unnecessary.

Animal suffering isn't pretend We evolved to think it is

Now that I've acquired a much bigger audience via posting about AI, my genius move is to start posting about the most alienating thing I believe andymasley.substack.com/p/animal-suf...

YouTube video by 80,000 Hours Why Governments Can’t Get Anything Done and How We Can Fix It | Ian Dunt

My interview with @iandunt.bsky.social on how the UK government got to be so bad.

It's not individuals but bad incentives and bad systems that make effective governance impossible.

We run through the UK's many institutional absurdities and how to fix them.

www.youtube.com/watch?v=1Yfo...

remember, “chatgpt is destroyeing the environment” is factually wrong and serves as a narrow dopamine/attention-maxed distraction from actual climate impact

A cheat sheet for why using ChatGPT is not bad for the environment Arm yourself with knowledge

I'm glad somebody out there is brave enough to push back against the "personal ChatGPT usage is terrible for the environment" message andymasley.substack.com/p/a-cheat-sh...

Incidentally, this is why I worry about the dominant position of AI critics on this site that "AI doesn't work and is just a scam"

It absolutely does work, in good ways (controlled studies on education, medicine) and in ways that create risks & problems. Pretending AI isn't real is a big mistake.

I wouldn’t say there aren’t benefits to echo chambers, but the cost to your ability to know true things is really high, even if the people involved have good political values.

It’s wild to me how much of an own goal this is. I’ve seen academic publishers miss such obvious opportunities to promote books for a wide audience and sell them at reasonable prices.

The academic monograph is dying because (a) publishers don't market them or price them affordably; (b) non-peer-reviewed trade books (because marketed and affordable) are more highly cited and academically influential.

I fear the Bluesky audience is not read to hear this😓

A cheat sheet for conversations about ChatGPT and the environment Arm yourself with knowledge

I wrote a cheat sheet version of my Using ChatGPT is not bad for the environment post. It's paired down and more focused on simple responses to common objections, without a long intro or my background environmental philosophy. andymasley.substack.com/p/a-cheat-sh...

We spent an absurd amount of time in 2022-24 denying this gap b/w consumer sentiment & behavior existed & inventing complex explanations like “felt inflation,” when the answers were right there. It was a sobering lesson in many things, including the politics of Narratives™️

bsky.app/profile/mcop...

Reality check: using ChatGPT and other LLMs is not a particularly bad use of time for the environment compared to other things you might do.

Lots of people who do have deep knowledge of AI consider these to be open questions. Dario Amodei, for one — and many others.

It’s hubris to assume the answers are obvious.

I don’t understand what people are mad about here

My theory: the relative decline in power of a certain kinda woke humanities set (post-2020) happened in some part because when a crisis came and America was relatively receptive they just had no workable policies. "Abolish the police" just weren't it, but it was the only slogan that broke through...

Thinking about the Nucleic Acid Observatory, AI will likely speed up some aspects of the creation of a detection system more than others. To the extent that we expect rapid advances in AI we should prioritize the work that we expect to bottleneck our future AI-accelerated work.

This is true. I find the dimethyl sulfide report intriguing, but it’s a big universe out there with a lot of chemistry going on in it. We just don’t know enough to say.

New preprint finding that eliminating US global health funding over the next fifteen years would cause:
- 15.2m deaths from AIDS
- 2.2m deaths from TB
- 7.9 additional child deaths

It's genuinely hard to express to sort of normal not politically plugged in non-lawyers *how* lawless the Trump administration is, and I think that's some part of the problem.

You sound like you're hyperventilating, but it's just a flat description of what is going on.

I think this is basically right. But many arguments against such ontological entities are able to meet their burden of proof.

AI 2027 A research-backed AI scenario forecast.

There has been a lack of near-term "hard science fiction" on might happen if supersmart AI is actually possible soon

This is just such a story written by some long-time key thinkers in AI. It is, no doubt, very wrong on most things, but reading it (& the footnotes) can still be useful. ai-2027.com

A picture of Henry George with a quote that says “protectionism teaches us to do to ourselves in time of peace what enemies seek to do to us in time of war”

Once again: reject McKinley, embrace George

A line graph titled "Confirmed cases of influenza B Yamagata" shows weekly confirmed cases of B Yamagata influenza worldwide from various types of influenza surveillance. The vertical axis represents the number of cases, ranging from zero to 5,000, while the horizontal axis covers the years from 2013 to 2025. The graph indicates significant fluctuations in case numbers over the years, with a noticeable peak just below 5,000 cases around 2019, followed by a steep decline. A data source note mentions that only a fraction of potential influenza cases is tested by labs for confirmation and strain identification. The data comes from FluNet, World Health Organization, and is dated 2023. The graph is licensed under CC BY.

A flu lineage has likely gone extinct since 2020 🧵

What would a world without foreign aid look like? The brutal debate over how to rescue foreign aid.

Found this from @dylanmatt.bsky.social very helpful: actually answered the basic questions about the USAID situation (which has been IMO weirdly hard to find in a lot of other reporting on this). Reminded me of the original need for Vox! www.vox.com/future-perf...