If you are completely unfamiliar with the actual science on obesity you probably think that's dumb because obesity is caused by high-palatability foods. Read the first page linked if you'd prefer to know why that's obviously wrong.
I admit to being, at present, persuaded by the high-palatability hypothesis, which I roughly translate into the following thesis: "The general rise in obesity is primarily explained by the rise of highly processed, addicting foods, which raises our natural set point, tricking our bodies into eating more calories than we 'need' before feeling full."
I read the posts you linked (you referred to this one, right?), and I'm not convinced by them, but I'm open to people explaining why they think I'm still wrong.
First I'll summarize the article briefly, and then respond to each point.
My brief summary
The series begins by outlining 8 mysteries:
- Obesity has gotten a lot worse over time
- Obesity abruptly got worse some time in the 1970s
- There's good evidence that we're not winning the war against obesity
- Hunter-gatherers don't become obese
- Lab animals and wild animals have also become obese over time
- People and animals gain a lot of weight when exposed to palatable fo
Impression: One of the articles also made a point about how certain tribes started becoming obese when exposed to Western culture. The high-palatability food hypothesis explains this reasonably well, whereas I'd need to see more details to imagine how lithium poisoning could have happened through their water supply.
Long ago, when SSC had an article about the altitude/obesity thing, a friend and I looked more closely at the data. I concluded that it seems like the bulk of the effect is explainable by selection effect, since there are very few people who live above a few thousand feet elevation, and they're probably disproportionately upper class and active. See https://slatestarcodex.com/2016/12/11/open-thread-64-5/#comment-443619 (and the original post at https://slatestarcodex.com/2016/12/05/thin-air/). I'm serious about these selection effects--the data linked in my comment includes BMI values up to 3km or 9,800 feet above sea level. I don't think there are 10,000 Americans living at that elevation total, and they almost all live in towns that primarily exist to serve wilderness recreation.
When Scott more recently posted about this hypothesis in one of the ACX open threads, one of the SMTM authors answered some questions in the comments. The mechanism tying elevation to pollution is allegedly that elevation is a proxy for how upstream you are in the water cycle, since water will accumulate toxins from the ground or being pumped into the water as it goes. To me, this seems like an extr...
The mechanism tying elevation to pollution is allegedly that elevation is a proxy for how upstream you are in the water cycle, since water will accumulate toxins from the ground or being pumped into the water as it goes. To me, this seems like an extremely loose association.
I agree, but moreover it looks like it should be an easy theory to test. My guess is that there are basically three routes for contaminants to enter our body and make us fat. The chemicals could be in the air, the water, or our food. If the SMTM authors believe that it's in our water, then drinking distilled or purified water should make us thinner. Do we have any evidence of this?
When I looked into it, you could see an effect on birthweight for babies born to mothers in high altitudes vs their lower-altitude siblings, and vice versa, which suggests to me something non-genetic is going on. And the effect of altitude on birth weight held up in countries where altitude was associated with both lower and higher income (although that wasn't the sibling study), which pushes against and doesn't eliminate income effects.
(Actual reality advisement, do not read if you'd rather not live in actual reality: Things really worth doing, thus in AGI alignment, are hard to come by; MIRI is bottlenecked more on ideas worth pursuing and people who can pursue them, than on funding, at this point. I think that under these conditions it does make sense for EA to ever spend money on anything else. Furthermore, EA does in fact seem bound and determined to spend money on anything else. I therefore think it's fine for this post to pretend like anything else matters; much of EA with lots of available funding does assume that premise, so why not derive valid conclusions from that hypothetical and go ask where to pick up lots of QALYs cheap.)
MIRI is bottlenecked more on ideas worth pursuing and people who can pursue them, than on funding
Ideas come from (new) people, and you mentioned seed planting which should contribute to having such people in 4-6 years, seems like still a worthy thing to do for AGI if anything is worth doing for any cause at all (given your short timelines). If you agree what's the bottleneck for that effort?
If ideas are the bottleneck, perhaps there should be monthly pitching sessions where people get to pitch their AI safety ideas to MIRI? Obviously, I imagine that you'd need to find someone who'd be a good filter - to ensure that there are only a reasonable number of pitches - whilst not cutting out any "crazy" ideas which are actually worth considering.
Well, this has inspired me to finally make an account for LessWrong commenting!
As someone very familiar with the "science on obesity", I do not find SMTM remotely persuasive. https://basedprof.substack.com/p/smtm-mysteries canvasses the reasons why in detail, for those who are going to look at the link in the above article and be inclined to trust it. SMTM repeatedly misrepresent not just scientific consensus, but also literally the articles which they cite. There are multiple examples of it in the "Mysteries" post, but this isn't even the worst offender.
If you do not believe that EA should be in the business of giving money to proven liars, shown to be rather specifically lying about the thesis which they would be given money to study, you should not give a cent to them, and you should point out that others should (not) do the same.
For those who don't want to read my (admittedly ranty) post about them (the reasons for said rantiness are, well, the consistent lying), here are three such lies/misrepresentations (two from the above, another from the second post in the series). As an aside, I'm not merely asserting that they are often wrong (though that would also, b...
R.e. #2
They provide literally no evidence for the proposition that the Hadza eat as much sugar as Americans do, their citations certainly do not demonstrate this.
The cited paper has a chart showing that Hadza get 14% of their calories from honey. A quick google search claims honey has 17g sugar per 64 calorie serving. So assuming 2000 calories/day, that's 280 calories of honey, which contains 74g sugar.
According to some google results, Americans consume around 70 (highest I saw was 77) grams of total sugar per day.
So it does seem to be similar. SMTM writes that this is "Combined with all the sugar they get from eating fruit". The same paper says 19% of their calories are from berries and another 18% from baobab fruit.
So it seems entirely plausible that SMTM is correct here.
I just read your post, and it seems like you're arguing against something SMTM don't support. They admit that they describe it badly in their initial post, but right after the CICO section there's a link to this additional post about CICO.
My understanding is that they're arguing two things: against the "linear relationship between calorie intake and weight change" meaning of CICO, and that it's mysterious that people are eating more over time:
...The overfeeding studies provide extremely strong evidence against this version of CICO [the linear relationship], since people gain very different amounts when overfed by the same amount, the difference appears to be mostly genetic, and some people actually lose weight, even when overfed by moderate (1000 kcal/day) amounts. Many people still believe something like “for every extra 3500 calories you eat you always gain one pound”, but all available evidence comes down very strongly against that.
At low levels of overfeeding, people at normal weight often don’t gain any weight at all.
[...]
[...] A common interpretation tied up in CICO is that differences in willpower explain the difference between obese and lean people. The idea is that weight g
(1) "their [our grandparents/great grandparents and older] diets were worse, not better. They ate more bread and almost four times more butter than we do today. They also consumed more cream, milk, and lard." - and the linked article once more: https://academic.oup.com/ajcn/article/91/5/1530S/4597478 I suppose this is not "strictly" a lie in the same way that "When did you stop beating your wife?" is not strictly an unreasonable question to ask. But it's deeply misleading, and SMTM must know it is misleading if they actually read the article they use as evidence. That very article cites a massive increase in consumption of other fats, and carbohydrates- so even if our ancestors ate more butter and bread, they were still consuming far fewer calories. This, in fact, is what the article concludes: "The increase in childhood obesity mainly reflects increased energy intake".
For those who would like to check this quickly, the data is on page 2 of this pdf.
I concur that the quoted statement is disingenuous albeit literally true.
Jacob Lagerros got excited about RaDVaC early and came to the Lightcone team a few months ago saying we should arrange funding for them. I believe talks have occurred between relevant parties though I'm not certain of exactly where things ended up.
Summary: I suspect funding has been provided, mostly want to give credit to those who noticed early and worked to make it happen.
I found a low six figure donor who filled 25% of their funding gap. The rest has not been filled. EDIT: They also got an ACX grant, but not enough to fill the whole funding gap I believe. I can intro donors who want to fill it to the radvac team. Email me at jacob at lesswrong.com
Original thread and credit to ChristianKl who picked up on and alerted me to rumours of the funding gap: https://www.lesswrong.com/posts/fBGzge5i4hfbaQZWy/usd1000-bounty-how-effective-are-marginal-vaccine-doses?commentId=XwA8mtvK8YCob2pLK#comments
I'm happy to hear that my cries resulted in at least some founding being moved to them even if it's still sad that there's still a funding gap.
In addition to what Eliezer already said about RaDVaC being good because it uses technology that can easily scaled up, RaDVaC has the potential to reduce infection by COVID-19 when given in addition to traditional vaccines because it potentially creates a mucosal immune system response in the upper respiratory system which the vaccines we currently have don't.
While there's some potential that Omicron will end COVID-19 that is optimistic. We might need something like RaDVaC to finally end this pandemic.
Hi all, this is Alex and Preston from RaDVaC. Thank you Eliezer for your thoughtful words of support. And thank you again Jacob L, your ongoing support is very much appreciated. We're also thrilled about receiving an ACX grant; not only is the money enabling, but the added recognition has already been extremely positive, and we have received additional donations following the award announcement. As Eliezer noted, we still don’t have all the funding we need to accelerate badly needed change in vaccine access and independence, but we do have funds to make great progress this year in ongoing projects.
One key goal is to establish an iterative cycle of vaccine improvement and testing, which will depend on improving our ability to measure immune responses. Another major goal for this year is to move toward full vaccine production independence, especially of vaccine antigens that are likely to be resilient against ongoing mutations. Vaccines based on recombinant antigen proteins produced in yeast are increasingly common, and we are in the early stages of designing yeast antigen factories that can be stored, shipped, and used as a foundational component for rapidly deploying vaccine factor...
Starting with early generations of RaDVaC vaccine designs, some of our core group engaged in self-experimentation to assess safety, as well as testing of some immuno-efficacy biomarkers. Four of us took a series of blood, saliva, & nasal wash samples over several months to determine antibody response in each. A few others also donated samples, which we also tested. We performed early antibody ELISA testing in our lab with custom tests. In parallel with our own testing efforts, we also established collaborations with academic researchers to perform antibody testing and neutralization assays on our samples, and we connected with colleagues in industry to perform T cell testing of various kinds.
In our lab, some of the data were encouraging but some results were conflicting, and overall, they were irreproducible. At the time we were self funded and couldn't afford more mistakes. Collaborators in both academia and industry were instructed by management and legal counsel not to test our samples. Even though our samples were obtained with full disclosure and individual consent, lawyers for otherwise collaborating institutions declared our project untouchable due to the absence o...
Thanks for the detail. I think that means that the answers to my questions are "no, not at all, for the more recent versions of RaDVaC, and in fact for earlier versions what we have is evidence that it usually doesn't produce any immune response", and "no, not at all, for any version of RaDVaC".
For the avoidance of doubt, I'm not saying that this means you're bad people, or that it means your thing doesn't work, or that it means no one should give you money. I'm just trying to assess how likely we should think it is that what you're doing will turn out to be useful.
I think, in particular, that when Eliezer (in the OP) says
They've pretty much proved themselves. They should be nurtured and scaled up to where they can start to replace US and Earth defunct state capacity to do the R&D that leads up to being able to rapidly design new vaccines that rapidly scale in production and deployment.
that first sentence is 100% false unless "proved themselves" means something like "proved that they are smart people" or "proved that they are acting in good faith" rather than "proved that they have something that works", which means that the second is waaaay premature.
Which, again, doesn'...
Following up to affirm two things:
Actoverco's data were informative about (positive) safety and immunogenic response in RaDVaC's Gen 8+9 designs. These data aren't immediately straightforward (antibody response was greater in immunologically "primed" individuals than in immunologically naive individuals), but really informative along the line of antigen selection, presentation, and dosing.
We (writ large) need vaccines that can be deployed rapidly and are focused on that rapid accessibility. There's no question that safe & effective vaccines can be made. The question for us is how to make safe & effective vaccines rapidly & simply enough to be widely accessible. Our approach has been to create and share designs openly, that can be (re)produced, adapted, tested, and studied without restriction. We couldn't do every stage of development in our own lab, so we invited others to use their expertise to contribute. Several have and still are, but I agree that it's time for us at RaDVaC to invest more (and more directly) in producing data that we can guarantee be shared openly.
We're working with outside people and groups to create appropriate trialing models and protocols, including a focus on challenge trials. We're looking to hire at least 2 consultants for the process (medical writing, trial design experts) and at least 2 onsite medical staff with fluency in the regulatory & logistics context of wherever a trial is to be run. (Depending upon location, we may also need an IRB's services to review and approve any trial design; definitely necessary for anything done within the US). One of the groups we are working with, 1DaySooner, has already created and published trial designs, and our trial design will be different but informed by the excellent foundation they've provided. We estimate that this preparatory phase will cost $500k-$1M.
An order-of-magnitude estimate for the full cost of a completed trial would depend on scale and location, but the cost for a relatively small but sufficiently powered trial in our current focus location of the Bahamas is in the $10M range.
There's an easier study I'd like to do before the lithium experiment: compare water contamination to obesity rates. I have two decent but not amazing datasets to do this (the water one tracks 18 contaminants but not lithium by zip code, the weight one tracks % obese not BMI by county) and a statistician who will analyze the results if I get the data in a single spreadsheet. I expect there are APIs to do that fairly easily but haven't had time to dig into it myself. If someone gets the 18 contaminants plus % obese in a single spreadsheet, and fixes the issues with county vs. zipcode aggregation, I can get this analysis done fairly quickly. Bonus points if you find better DBs than I do (I wish it was BMI by zipcode, not % obese by county) or incorporate additional data, like income, age, well water usage, and density.
I've discussed this with SMTM previously, they agree it's worth doing although are less excited than me because the water quality database doesn't include lithium. So additional bonus points if someone finds a database of lithium concentration in drinking water by county.
I'd happily pay $100 + credit on the eventual blog post for the spreadsheet combining the two DBs listed, and am open to negotiation on larger amounts if it's more difficult than I think it is or additional features are included.
One of the mysteries mentioned in the SMTM article was that the obesity epidemic began abruptly some time around 1980. The authors state,
This wasn’t a steady, gentle trend as food got better, or diets got worse. People had access to plenty of delicious, high-calorie foods back in 1965. Doritos were invented in 1966, Twinkies in 1930, Oreos in 1912, and Coca-Cola all the way back in 1886. So what changed in 1980?
This article argues (convincingly, in my opinion) that the obesity epidemic began much earlier, probably around the turn of the 20th century. If you track the evolution of BMI over time in different birth cohorts, you get a very continuous, smooth rise in obesity.
Totally agree! I stumbled upon the SMTM link 6 months ago or so and it was a big view-changer for me. I'd previously thought calories-in-calories-out was the main thing to be focusing on but, uh, yeah, I was super wrong.
I do think you're too critical of the high-palatability theory. The SMTM page finds the theory reasonable:
...“Palatable human food is the most effective way to cause a normal rat to spontaneously overeat and become obese,” says neuroscientist Stephan Guyenet in The Hungry Brain, “and its fattening effect cannot be attributed solely to its fat or sugar content.”
Rodents eating diets that are only high in fat or only high in carbohydrates don’t gain nearly as much weight as rodents eating the cafeteria diet. And this isn’t limited to lab rats. Raccoons and monkeys quickly grow fat on human food as well.
We see a similar pattern of results in humans. With access to lots of calorie-dense, tasty foods, people reliably overeat and rapidly gain weight. But again, it’s not just the contents. For some reason, eating more fat or sugar by itself isn’t as fattening as the cafeteria diet. Why is “palatable human food” so much worse for your waistline than its fat and sugar alon
Isn't lithium in water linked to lower depression rates and not really something you'd want to straightforwardly remove even if it turned out to be making people fat? I guess you might win on net if it turned out you could cure about that much depression with lotsalightboxes and be rid of obesity in the bargain, but it's at least a little complicated.
My review mostly concerns the SMTM's A Chemical Hunger part of this review. RaDVaC was interesting if not particularly useful, but SMTM's series has been noted by many commenters to be a strange theory, possibly damaging, and there were, as of my last check, no response by SMTM to the various rebuttals.
It does not behoove rationalism to have members that do not respond to critical looks at their theories. They stand to do a lot of damage and cost a lot of lives if taken seriously.
For those who'd rather read that, the author of the obesity articles series summarized them in a twitter thread (Thread Reader). There's also discussion of it on LW (though it took place before most of the articles were published)
Having not apparently the energy to write this longly, I write it shortly instead, that it be written at all.
Just a comment on writing for understandability — compare Benjamin Franklin, writing in 1750:
I have already made this paper too long, for which I must crave pardon, not having now time to make it shorter.
Shorter is (almost always) better, please don't write things longly just for the sake of it!
Other types of causes I would like to see more groups working on and more people supporting are those that would help make human communities more robust against civilizational collapse (minus scenarios where we're all turned into paperclips, of course). Right now, billions of humans are utterly dependent on global economic infrastructure to supply food, water, energy, shelter, etc. If some event breaks down this infrastructure, billions could die, since not only do most people lack survival skills, but the local resources in most areas of high population density are not sufficient to provide enough for (even a small fraction of) everyone. Ideally, every local community, from small villages to sprawling metropolises, could become locally self-sustaining to the point where getting cut off from the global economy would lead to the loss of luxury items and foods rather than to mass starvation.
Things that come to mind include the Global Village Construction Set, an open-source set of 50 blueprints for technologies explicitly designed to require minimal material and manufacturing resources to construct but that could be used to rebuild civilization in the event of collapse.
Food productio...
I like the shortlyness of this post and think it perfectly suits the subject. For many purposes longlyness can be counterproductive.
Is there any reason we can't quickly test the lithium hypothesis by buying a bunch of random food and paying a lab to grind it up and figure out how much lithium is in it (use a mass spec?)? I assume this would cost thousands of dollars, but that's pretty cheap.
I thought it was a mix:
- Lithium from drinking water
- Lithium concentrated in certain foods (see Wolfberry section in https://slimemoldtimemold.com/2021/10/11/a-chemical-hunger-interlude-g-li/ for example)
Fort Collins, CO gets water almost exclusively from rivers starting nearby and still has an 18% obesity rate (and Boulder is similar but has 12%). Both of those are "low" compared to other places, but still way above the historical level.
(I admit that I'm less interested in the water hypothesis since I live in Fort Collins so that part is already solved for me)
This would be bad news for the tourism industry in Banff. Below was taken on my honeymoon.
To boot:
...Not only did tourists soak in the spring water, they drank it, too. Bob Elliott, acting operations manager at Banff Upper Hot Springs, said, “Local bars sold bottled hot springs water as a tonic and cure for hangovers. Banff's mineral water was bottled and sold as ‚Lithia water' for a short time
I’m looking at getting into grant writing; would that be something worth applying for grants (heh) for?
From what I have heard (I have not researched any of this very thoroughly), the palatability is not the problem directly but something very related is. It is not the case that someone would eat a lot, just because it is just so tasty. It is rather about that the composition of processed food is often very different from unprocessed food. And this affects how our body responds, like when we feel full. Eating some Froot Loops is very different from eating a mango.
This might not be the only, or even the main effect, but I would guess that it is a significant ...
If you ever want an apple tree that bears fruit, you may need to at some point plant a seed
You're going to need to plant more than one.
Having not apparently the energy to write this longly, I write it shortly instead, that it be written at all.
It's nice to see a decent post that's short.
UPDATE: Be advised that I no longer consider SMTM to be among the things we should be doing, though I don't regret having given them an earlier try. I wish somebody else was running out and trying the sort of things that SMTM is trying, but they're bad enough at analysis to not qualify for further support until that gets fixed. See eg this Twitter thread.
Having not apparently the energy to write this longly, I write it shortly instead, that it be written at all.
People sometimes go about saying now, in this community, that there is collectively enough money that we could potentially go do more things with it, if there were things worth doing.
If that's true, I'd like to see us planting seed grains now for replacing the now-defunct state capacity of the USA and Earth, with respect to biodefense and ultra-high-leverage med R&D.
As a poster child of a previous intervention I backed here, Seasonal Affective Disorder affects 0.5%-3% of the population and higher in countries at extreme latitudes. Call it a hundred million people at a guess. Standard lightboxes don't work very well for treating it. The Sun works great for treating it. Reason suggests trying more light. Earth, however, previously lacked the state capacity to investigate this obvious question of massive scope. I put a private funder in touch with a group that did a preliminary investigation; they need to run a larger study but preliminary results suggest that they may, in fact, need to use more light and that doing so might be effective. Again, that's for a problem that probably over a hundred million people have. So it's not very surprising that investigation of it bottlenecks on there being a single interested medical researcher, who writes Eliezer Yudkowsky as the person who posed that question, and gets funding arranged by that route. It was less than $100,000.
This sort of thing seems obviously competitive to me with QALYs/$ or DALYs/$ on global poverty interventions.
Really obviously competitive, actually. I will not argue the point further because I expect that most readers who can be persuaded found a glance at the numbers sufficient.
EA's ability to do more things in this space, if it is not bottlenecked on money, is very likely bottlenecked on: good ideas to pursue; people who can pursue those ideas; and/or admin staff who can investigate grants, make them, and operate them.
I suspect that people who can do these things are, at the very least, valuable. They should therefore be nurtured even in their earlier stages. If you ever want an apple tree that bears fruit, you may need to at some point plant a seed that doesn't have any apples on it right then, and invest effort into watering it in proportion to how many apples you hope for later, rather than how much this tiny sapling has proven itself right now.
I think it was a huge, huge mistake that more money was not spent on AGI alignment when it was small and weird and unproven. The resulting damage was not something that could be fixed by any or all of the money that became available later. Someday I may write more about this. Anyways, don't do that.
The Rapid Deployment Vaccine Collaborative, aka RaDVaC, is now past the small weird stage. They've pretty much proved themselves. They should be nurtured and scaled up to where they can start to replace US and Earth defunct state capacity to do the R&D that leads up to being able to rapidly design new vaccines that rapidly scale in production and deployment. If we can't give RaDVaC $2M I'd like to know what it's being spent on that's more important. Covid-19 was not very much of a pandemic and if Earth ran into a serious pandemic in its current state that would be a serious problem.
Some weird person [correction: 2 weird people] wants to investigate whether the real driver of our massive planetwide obesity epidemic is lithium contamination, probably in water. If you are completely unfamiliar with the actual science on obesity you probably think that's dumb because obesity is caused by high-palatability foods. Read the first page linked if you'd prefer to know why that's obviously wrong. If you know about the actual epidemiology of obesity and how ridiculous it makes the gluttony theory look, you are still probably saying "Wait, lithium?" This is still mostly my own reaction, honestly. But obesity remains a massive growing planetary problem, and almost nobody is investigating it in a way that takes the epidemiological facts seriously and elevates those above moralistic gluttony theories. If some weird person wants to go investigate, I think money should be thrown at them, both to check the low-probability massive-high-value gamble, and also to encourage them to maybe go investigate something else in the same space if the lithium thing doesn't pan out. $200K maybe. They advertise their effort under the name of Slime Mold Time Mold (SMTM) for no particular reason I can discern. I think this is the kind of decision that some people make when they are small and weird, and I think that if we allow that to prevent us from throwing money at them then we are stupid and not learning from history.
That's all. Writing this up now, briefly and hurriedly, because the
lithium person wantslithium people want to apply for some EA grants. Consider this my recommendation thereof.