Thanks for doing this check!

That’s 9.5% of the kids they caused to be vaccinated, which already seems highish - GiveWell estimates about 1 in 10 of these kids would have died from a vaccine preventable illness before age 5 if not for the New Incentives program.

This seems very different from what they claim in their spreadsheet (300 under-5 deaths averted per 10,000 children counterfactually vaccinated, so 3%)

Do you have a sense of what's driving the discrepancy?

Bloomberg's valuation of Moskovitz's fortune recently dropped by ~60% (from $30B to $11B) as his level of ownership of Meta was not significant enough to show up in their filings.

(But Forbes' estimate didn't change much at $19B)

Depopulation is Bad

Even the most pessimistic estimates expect the global population to actually start declining only in ~2050 and to reach the 1960 population levels well after 2100.

If this actually becomes an issue by 2100, I'm optimistic it will be relatively tractable by then

Every now and then I'm reminded of this comment from a few years ago: "One person's Value Drift is another person's Bayesian Updating"

GiveWell also takes outside donors and OpenPhil doesn't.

I don't think that's true anymore: https://www.openphilanthropy.org/partner-with-us/ but I imagine OpenPhil only takes donors above a certain size (here they say >$1M/year) while GiveWell takes donations of all sizes

Why have 2 different bodies doing the same thing with largely the same pool of money?

It doesn't apply to the TSU grant, but note that a high percentage of GiveWell-directed donations don't come from OpenPhilanthropy:

And I expect this to increasingly be the case in the future, as GiveWell finds new donors and OpenPhil finds other things to donate to. So I wouldn't say it's "largely the same pool of money"

Related to this point, I was surprised to see this

Given that GiveWell's All Grants Fund has basically the same graph

Many other grants from the All Grants Fund don't have a ton of evidence behind them and are exploratory. As an example, they funded part of an RCT on building trailbridges in Rwanda, with reasoning «While our best guess is that bridges are below the range of cost-effectiveness of programs we would recommend funding, we think there’s a reasonable chance the findings of the RCT update us toward believing this program is above our bar. [...]» and an RCT on providing eyeglasses for similar reasons.

newcomers discover that most funding and opportunities actually focus on longtermism (especially biosecurity and AI)

Another comment mentions GiveWell (moving $200M/year), but I want to also flag that Founders Pledge just announced receiving a $50M donation to their climate fund, operating at $100M+ scale. Giving Green received $10M last year, and estimated the same donor may actually have given $17 million more [to recommended climate change orgs]

This is smaller than GiveWell (for now at least) but more than what OpenPhil donates to biosecurity (~$35M/year). For context here are some numbers from OpenPhil's grants database (sheet)

There seems to be a lot of money in EA Enviromentalism

I think this is very hard to predict, and I just feel uncertain. Public perception seems to be really fickle, and I could imagine each show being either:

1. Negative towards SBF/Caroline and negative towards EA (it's all tech bros feeling superior, e.g. here)
2. Negative towards SBF/Caroline and positive towards EA (they used ethics as "mostly a front" and only cared about winning)
3. Positive towards SBF/Caroline, and negative towards EA (they started as idealistic altruists and got corrupted by the toxic EA ideology)
4. Positive towards SBF/Caroline, and positive towards EA (e.g. making John J. Ray III and Sullivan and Cromwell the villains)

And for each of these 4, it's not clear what the impact on EA would be, e.g. I think "The Wolf of Wall Street" probably got many people excited about working in finance.

I predict the documentaries will be negative towards EA, as was the vast majority of media on EA in 2023 and 2024, and I think documentaries tend to be mostly negative about their subject, but I'm much more unsure about the fiction series

Besides the selection bias mentioned in the post and other comments, I'm also unsure about how much to update on people's assessment of the mental health effects of their actions. I expect most people overestimate the positive mental health effects of any activity they choose to do.

As an extreme example, I expect that most smokers think that smoking is neutral or positive for their mental health, despite probably being net negative.^[1]

1. ^{^}

   I don't know what the research says about this, but LLMs seem to agree:
   
   ChatGPT Deep Research

   

   
   Gemini Deep Research

   

    

   Claude Research

   

    

    

   
   Smoking is an extreme comparison, but the same happens with e.g. veganism
   
   ChatGPT

   

    

   Claude

   

    

   Gemini didn't only consider perceived effects, so it's less relevant