This research piece is not as rigorous or polished as usual. I wrote it quickly in a stream-of-consciousness style, which means it’s more reflective of my actual reasoning process.

My understanding of HIIT (high-intensity interval training) as of a week ago:

1. VO2max is the best fitness indicator for predicting health and longevity.
2. HIIT, especially long-duration intervals (4+ minutes), is the best way to improve VO2max.
3. Intervals should be done at the maximum sustainable intensity.

I now believe those are all probably wrong.

Continue reading

An if-then commitment is a framework for responding to AI risk: “If an AI model has capability X, then AI development/deployment must be halted until mitigations Y are put in place.”

As an extension of this approach, we should consider retroactive if-then commitments. We should behave as if we wrote if-then commitments a few years ago, and we should commit to implementing whatever mitigations we would have committed to back then.

Imagine how an if-then commitment might have been written in 2020:

Pause AI development and figure out mitigations if:

• AI exhibits what looks like deceptive or misaligned behavior, or feigns alignment (1, 1b, 2)
• AI breaks out of containment in a toy example
• AI finds a real-world zero-day vulnerability
• AI qualifies for Mensa¹
• AI exhibits some degree of agentic capabilities
• AI writes malware

Well, AI models have now done or nearly-done all of those things.

We don’t know what mitigations are appropriate, so AI companies should pause development until (at a minimum) AI safety researchers agree on what mitigations are warranted, and those mitigations are then fully implemented.

(You could argue about whether AI really hit those capability milestones, but that doesn’t particularly matter. You need to pause and/or restrict development of an AI system when it looks potentially dangerous, not definitely dangerous.)

Notes

Conventional wisdom says charity cost-effectiveness obeys a power law. To my knowledge, this hypothesis has never been properly tested.¹ So I tested it and it turns out to be true.

(Maybe. Cost-effectiveness might also be log-normally distributed.)

• Cost-effectiveness estimates for global health interventions (from DCP3) fit a power law (a.k.a. Pareto distribution) with \(\alpha = 1.11\). [More]
• Simulations indicate that the true underlying distribution has a thinner tail than the empirically observed distribution. [More]

Continue reading

Summary

Last updated 2024-11-20.

It’s been a while since I last put serious thought into where to donate. Well I’m putting thought into it this year and I’m changing my mind on some things.

I now put more priority on existential risk (especially AI risk), and less on animal welfare and global priorities research. I believe I previously gave too little consideration to x-risk for emotional reasons, and I’ve managed to reason myself out of those emotions.

Within x-risk:

• AI is the most important source of risk.
• There is a disturbingly high probability that alignment research won’t solve alignment by the time superintelligent AI arrives. Policy work seems more promising.
• Specifically, I am most optimistic about policy advocacy for government regulation to pause/slow down AI development.

In the rest of this post, I will explain:

1. Why I prioritize x-risk over animal-focused longtermist work and global priorities research.
2. Why I prioritize AI policy over AI alignment research.
3. My beliefs about what kinds of policy work are best.

Then I provide a list of organizations working on AI policy and my evaluation of each of them, and where I plan to donate.

Cross-posted to the Effective Altruism Forum.

Continue reading

Last updated 2024-10-23.

Outlive: The Science & Art of Longevity by Peter Attia (with Bill Gifford¹) gives Attia’s prescription on how to live longer and stay healthy into old age. In this post, I critically review some of the book’s scientific claims that stood out to me.

This is not a comprehensive review. I didn’t review assertions that I was pretty sure were true (ex: VO2 max improves longevity), or that were hard for me to evaluate (ex: the mechanics of how LDL cholesterol functions in the body), or that I didn’t care about (ex: sleep deprivation impairs one’s ability to identify facial expressions).

First, some general notes:

• I have no expertise on any of the subjects in this post. I evaluated claims by doing shallow readings of relevant scientific literature, especially meta-analyses.
• There is a spectrum between two ways of being wrong: “pop science book pushes a flashy attention-grabbing thesis with little regard for truth” to “careful truth-seeking author isn’t infallible”. Outlive makes it 75% of the way to the latter.
• If I wrote a book that covered this many entirely different scientific fields, I would get a lot more things wrong than Outlive did. (I probably get a lot of things wrong in this post.)
• When making my assessments, I give numeric credences and also use terms such as “true” and “likely true”. The numbers give my all-things-considered subjective credences, and the qualitative terms give my interpretation of the strength of the empirical evidence. For example, if the scientific evidence suggests that a claim is 75% likely and I understand the evidence well, then I rate the claim as “likely true”. If I only read the abstract of a single meta-analysis, and the abstract unequivocally supports the claim but I’m only 75% sure that the meta-analysis can be trusted, then I rate it as “true”. Both claims receive a 75% credence.

Now let’s have a look at some claims from Outlive, broken down into four categories: disease, exercise, nutrition, and sleep.

Continue reading

Update 2025-01-17: I discovered another protein quality calculator that’s much more comprehensive than mine: https://www.diaas-calculator.com/

You may know that complete proteins are good because they contain every essential amino acid. But you might not know that that’s not the full story.

Take wheat. Wheat is a complete protein—it contains all nine essential amino acids. But it has a problem. Wheat only contains 27mg of lysine (an essential amino acid) per gram of protein, whereas the Food and Agriculture Organization recommends 48mg of lysine per gram. To make full use of a gram of protein, your body needs to get those 48mg. It doesn’t matter that wheat has lots of other essential amino acids. Once your body uses up all the lysine, it can’t make good use of the other amino acids in wheat protein.

You can evaluate the protein quality of a food using the Digestible Indispensable Amino Acid Score (DIAAS). This score determines the quality of a source of protein based on which essential amino acid will run out first, adjusted for digestibility. A score of 100 means the protein has plenty of every essential amino acid.

Sometimes you can improve the protein quality of your food by mixing different ingredients. Wheat has a DIAAS of 57 because it only has 57% as much lysine per gram as your body needs. Peas have a score of 82 because they don’t have enough methionine + cysteine. But peas have 131% of the lysine requirement, and wheat has 149% of methionine + cysteine, so mix them together and they cover for each other’s weaknesses. A 50/50 mixture of wheat and pea protein has a DIAAS of 94.

With this calculator, you can determine the DIAAS for mixtures of different protein sources.

Continue reading

You don’t always save money by putting your investments into a 401(k).

When you invest money inside a 401(k), you don’t have to pay taxes on any returns earned by your investments. But you also have to pay a fee to your 401(k) provider.

• If you buy and hold index funds in a taxable account, you don’t have to pay any capital gains tax on price increases until you sell.
• In a 401(k), the annual fee adds up every year and may eventually exceed the tax savings.

So the taxes cap out at the capital gains tax rate (15% or 20% depending on your tax bracket),¹ whereas the expenses of a 401(k) continue to accumulate.

However, in a taxable account, you do still have to pay taxes on dividends (and bond payouts) every year, and those taxes might cost you more than the 401(k) fees.²

Below is a calculator to determine how many years before the 401(k) fees exceed the tax savings, if ever.

employer matching (%)
total investment return including dividends (nominal) (%)
dividend yield (%)
401(k) fee (%)
capital gains tax rate (%)
income tax rate today (%)
income tax rate in retirement (%)

A 401(k) falls behind a taxable account after:

This calculator assumes you buy index funds and hold them forever. If you trade stocks within a taxable account, you have to pay taxes every time you make a trade.

Something else to consider: If you quit your job, your old employer’s 401(k) provider will let you roll your 401(k) into an IRA. You don’t have to pay any fees on an IRA.³ So even if the 401(k) fees exceed the tax benefits after (say) 30 years, that’s not a problem if you expect to quit your job after less than 30 years. Realistically, few people stay at one job for so long that the 401(k) fees exceed the tax savings.

(If you change jobs, usually you can roll your old 401(k) into your new 401(k), but I wouldn’t do that because it means you have to keep paying 401(k) fees. It’s almost always better to roll your old 401(k) into an IRA.)

Notes

I did another caffeine experiment on myself. This time I tested if I could have caffeine 4 days a week without getting habituated.

Last time, when I took caffeine 3 days a week, I didn’t get habituated but the results were weird. This time, with the more frequent dose, I still didn’t get habituated, and the results were weird again!

Continue reading

Last time, I found that the healthiest BMI range for all-cause mortality is 20–22. But BMI doesn’t tell the whole story. Most obviously, it doesn’t account for body fat vs. lean mass. All else equal, you’d rather have more muscle¹ and less fat.

So what’s the healthiest combination of lean mass + fat mass?

I’m not going to answer that question because I can’t. Instead, I will explain why I can’t, and then give a rough guess at the answer.

Scientists have been measuring and collecting data on BMI for decades. You can find plenty of giant BMI studies with three million participants in various countries.

We have much sparser data on body fat. Scientists didn’t start collecting data on body fat until the last few decades. And body fat is harder to measure—we have various methods for estimating body fat, but they’re all more complicated than calculating BMI.

I managed to scrounge together some studies on body fat and mortality. My best guess: the average woman should aim for a BMI of 21 with 20% body fat, and the average man a BMI of 21 with 10% body fat. (Subject to individual variation due to genetics and whatnot.)

Trans men should probably target the same body fat % as cis men, and likewise for trans women and cis women, because hormone therapy alters body fat distribution (Spanos et al. (2020)²).

The evidence weakly suggests that there is no lower bound on healthy fat mass, and no upper bound on healthy lean mass. We have so little mortality data on extremely lean + muscular people that we can’t say how healthy they are.

A more in-depth analysis would look at a variety of health indicators (blood pressure, HDL cholesterol, etc.) and use that to predict mortality. I didn’t do that, I just looked at mortality data.

Continue reading

TLDR: 20 to 22.

Continue reading