Oh, this is horrible, no joke:
What could go wrong with eating an extremely high-fat diet of beef, cheese, and sticks of butter? Well, for one thing, your cholesterol levels could reach such stratospheric levels that lipids start oozing from your blood vessels, forming yellowish nodules on your skin.
That was the disturbing case of a man in Florida who showed up at a Tampa hospital with a three-week history of painless, yellow eruptions on the palms of his hands, soles of his feet, and elbows. His case was published today in JAMA Cardiology.
The man, said to be in his 40s, told doctors that he had adopted a “carnivore diet” eight months prior. His diet included between 6 lbs and 9 lbs of cheese, sticks of butter, and daily hamburgers that had additional fat incorporated into them. Since taking on this brow-raising food plan, he claimed his weight dropped, his energy levels increased, and his “mental clarity” improved.
I’m uncomfortably reminded of my friend Seth Roberts, who died of heart failure not long after adopting a diet in which he ate half a stick of butter a day, which he claimed from self-experimentation to have improved his brain function.
The Florida man was ingesting a lot more than half a stick of butter per day; still, I was struck by his report of improved mental clarity, which uncomfortably echoed Seth’s claims from a decade earlier.
I feel some very small responsibility for this chain of events. Seth was a friend of mine at the University of California in the early 1990s–we taught a couple of courses together–and he kept me informed of his work on self-experimentation. In 2005 he published a research article on the topic, and I promoted it on the blog. Even then, his self-experimentation was unusual, but he hadn’t gone off the deep end. We had further blog discussion that year (also here), and Alex Tabarrok linked to it from his Marginal Revolution blog. From there it was picked up by Stephen Dubner and Steven Levitt in the New York Times. At the time I thought this multiple stage of amplification was pretty cool, and at this point Seth didn’t need my help to get publicity. Dubner and Levitt invited him to guest-blog on his diet at Freakonomics, then a year later Seth published a successful diet book, which I reviewed positively, as did Tabarrok. Around that time, I also published a couple of conversations with Seth on his research; see here and here. Seth also started his own blog and website where people using his diet could share tips and otherwise communicate. He was followed by mid-level celebrities such as Dennis Prager and Tucker Max.
Seth was an early figure in the paleo-lifestyle movement. Indeed, he was telling me his theories about the benefits of caveman diet and caveman lifestyle many years before I heard it anywhere else. In the late 1990s or early 2000s he even wrote a book on the topic, but he never found a publisher, and I guess he abandoned the project after publishing his diet book.
Going through the blog archives, I was surprised to learn that, as early as 2007, Seth was claiming cognitive benefits from oil consumption. He shared data showing it had improved his reaction time. I think this must have been a byproduct of him drinking oil as part of his weight-loss plan. Here’s what I wrote back in 2007:
Encouraged by the success of his self-experimentation to help his sleep, mood, and weight concerns, Seth Roberts has been experimenting with the effects of drinking flaxseed oil. . . . Commenting on another recent one of Seth’s self-experiments, I wrote,
Seth,
Not to be a wet blanket or anything, but aren’t you worried that your findings might be due to expectation effects: you knew which oil you were taking when doing the tests, right?Seth replied,
Andrew, no, I’m not worried that the results are due to expectations. If the results always conformed to my expectations, I’d be worried, but they haven’t — see my post about eggs. Moreover, this particular result confirms a result that was a surprise. In other words, I’ve gotten the same result when I was expecting it and when I wasn’t expecting it.
I’m still concerned, though. Seth is saying that it’s not just an expectation effect because he wasn’t always expecting the results. But I could see a bias arising from positive feedback, as follows: You try a new treatment and then see what happens after, with no expectations except that things might change. There is some noise to this measurement–just at random, it will be higher or lower than before. Having seen this, you adjust your expectations; this then affects your next measurement, etc.
In retrospect, I think I was too mild. I suggested to Seth: “maybe you could get a partner in experimentation, someone who lives or works nearby, and he or she could give you a randomly assigned oil. That is, your partner would know which oil you’re getting, but you wouldn’t. In fact, you wouldn’t even know if you were being given something new that day.” Seth agreed but wrote, “they are low on my to-do list.”
Again, I wish I’d pushed back harder on that:
– A clean negative result could’ve given Seth (and me!) a healthy dose of skepticism and motivated him to think more carefully about his theories when deciding what experiments to try next.
– A clean positive result could’ve given him and his followers a level of confidence and motivated him to experiment more systematically, rather than spiraling out of control with goofy and possibly dangerous ideas such as eating a half a stick of butter a day.
We had another exchange in 2009, where, again, I said, “maybe this could all be a confirmation bias,” and, again, Seth pushed back, and, again, I now wish I’d taken a firmer line.
As the years went on, Seth gained confidence and his experiments became more and more dubious, culminating in the stick-of-butter thing which, maybe it wasn’t fatal, but maybe it was, and I doubt it was helpful, either to his well-being or others’.
This is a long chain of events, so I don’t feel a high degree of responsibility for my small part in encouraging the crazier aspects of the paleo lifestyle movement. But, to the extent that I was proud that my promotion of Seth’s work led to his media success, I should correspondingly feel bad about the negative consequences.
In a world where it is hard to know who/what to trust, you relied on a friend who you had reason to trust. You queried him repeatedly. You considered revising your views repeatedly. Isn’t that all we can expect of ourselves? It’s ok to regret not having expressed stronger doubts, or reduced you involvement in promoting/supporting Seth’s claims, but if things had turned out differently would you have regretted not being more forceful in support?
For me, this is a good story about the best you can do in a world where it is hard to trust anything. If only current supporters of the Administration (thread hijack alert!) were to do the same. I see no pushback, no Bayesian updating, no regret. It’s fine to put your faith in someone you trust, but that is the beginning not the end of your questioning.
Dale:
I feel that I was ok in my interactions with Seth. In retrospect I should’ve had more skepticism from the start, but things were different thirty years ago. We’re all a lot more skeptical than we used to be. Where I feel bad is that after updating my views I didn’t push harder on the skepticism. In the first part of the story, my unskeptical or weakly skeptical promotion of Seth’s ideas was amplified through Marginal Revolution, then the New York Times, then Freakonomics, to reach a wider audience and to add fuel to the paleo diet movement. In the second phase, I expressed my skepticism here on the blog but it did not get amplified. One reason is that all I had was skepticism–I didn’t have a “debunking” to offer. Seth made large claims that I think were often absurd claims, but to his credit he did not overstate the evidence underlying his claims. All he ever did was give the reasoning behind his ideas and say that they observationally worked for him and others. So there was never going to be a slam-dunk refutation in the same way that was done with beauty-and-sex ratio, air rage, power pose, himmicanes, etc.
I’m not sure if you’re aware of just how popular these N=1, self-experimentation “health influencers” have become. There’s a lot of them.
Which, I think, suggests a kind of internal driver that isn’t particularly influenced by skeptical, external feedback. In fact, as seen with the weaknesses of the “information deficit model,” often times external pushback only reinforces the motivation to double down and dig in.
Yes, that stuck out to me too. There are now hundreds of thousands of people like this out there. Maybe the main difference is that a lot of them are chiefly grifters rather than the terminally curious, which appears to be the case with Seth.
Robin:
I don’t think that Seth was at all a grifter. That said, I do think he enjoyed the positive attention he got on the internet, so there was some incentive for him to make wilder and wilder pronouncements. He was in a bad feedback loop with his audience.
It’s not surprising that any prolonged extreme diet has bad consequences. But a lot of people benefit from a short term extreme diet.
Many have said that after going carnivore for a month they discovered had been intolerant of something they were eating, say gluten or some food additive. The benefit wasn’t from what they ate but from what they didn’t eat. They might have made the same discovery from going vegan for a month.
John:
Agreed. Self-experimentation can be useful. People can go on diets all on their own and learn things. That was one of the challenges of dealing with the Seth Roberts phenomenon, that he had a mix of reasonable and unreasonable ideas. In some ways it’s easier to talk about things like the UFOs-as-space-aliens thing, because that’s just crap through and through. Work that’s on the border is harder to think about.
The human body often benefits from mild shocks. The problem arises when attempts are made to consolidate these gains over a longer period. People who know they can’t follow an extreme diet forever invariably tend to lose some of the gains they made when they switch back to their regular lifestyle, even with the changes prompted by their experiment. This is because the only way to keep long term changes is to do the right things in the long term. These right things are generally measures that are only moderately better than their previous ways.
It is a case of overoptimization and of taking ideas and their implementation to their logical conclusion, a practice that doesn’t seem to work out well in complex systems.
I doubt 6-9 lbs (20-30 kcal, or 10x a normal diet) per day is typical for a “carnivore diet”.
This brings up a couple of points though.
1) Something like theranos is needed so people can cheaply test their various biomarker levels. The theranos tests failed to meet some arbitrary regulatory/certification level of accuracy but were perfectly fine as a screening test for gross deficiencies/overdoses.
2) There is this obsession in medicine of splitting/grouping things into heuristic categories (like gender, race, socioeconomic). But then everyone forgets thats all it is, just a convenient heuristic. The categorization needs to be applied intelligently.
Like in this case, you really think the relevant category is “carnivore diet” and not 30 kcal per day? He probably had some genetic susceptibility too.
Anon:
I appreciate a bit of contrarianism but this is ridiculous. Theranos was not “perfectly fine”–they were just ripping off existing tests that were produced by others. It would be like . . . ummm, suppose I bought a bunch of 1-ounce Hershey bars for $1 each, and then I put my own label on them saying they were 2-ounce bars of gourmet chocolate from my artisanal labs or whatever, and then I sold them for $2. What I’m doing then is lying about where the chocolate’s coming from, lying about how much chocolate I’m offering, and ripping people off. It’s hardly a defense to say, Hey, I’m not meeting some “arbitrary regulatory/certification level of accuracy” but it’s perfectly fine because I’m supplying chocolate to people. Take away the mislabeling, and I’m just offering a $1 product for $2. Similarly with Theranos. The tests they were ripping off already existed. The only value added by Theranos was that they were lying, and this helped them illegally and unethically raise tons of money.
If you want to share your contrarian takes on Theranos, this is not the place to do it. It’s just polluting our comment threads. If you want to talk about actual blood testing companies, that’s fine, and no need to mix that with some business fraud people.
Theres nothing contrarian about it. I simply looked at the data. It was brought up here:
https://statmodeling.stat.columbia.edu/2023/06/29/if-the-nobel-prize-winners-and-the-bigshots-at-the-national-cancer-institute-are-doing-hype-on-the-daily-then-it-makes-sense-that-the-loser-wannabes-wannabe-bigshots-at-the-cleveland-clinic-and-the-s/#comment-2233949
The way I would use these tests (sequentially) the less accurate results are fine. And after reading the responses, I still think so.
The one response that made sense was that diluting small samples and running standard assays (which is possibly what they did for some/all of them) would not be any cheaper.
But even in that case Id still often prefer getting less accurate results back for multiple samples per day over the course of a week than one very accurate snapshot.
Ignoring the dynamics for logistical reasons is a much bigger problem than the loss of accuracy reported there, imo.
Anon:
If it’s a good idea, I’m sure it’s being done by people who are not lying criminals.
Well I’m here, waiting, ready to pay for something like that.
Where do people who advocate a “paleolithic” diet sleep? Are they sleeping in an actual cave? I don’t think that there was much pastoralism 25,000 years ago. Butter and cheese did not exist. Hamburger supposedly was invented by Tartar tribesmen carrying chunks of meat in their saddlebags. Domestication of the horse and invention of saddlebags are not paleolithic either. Modern farm animals are much fatter than critters eaten by people a thousand years ago. Let’s just call it a high fat diet. You are free to devise any dietary hypothesis you want and even live accordingly, but please don’t invoke some imaginary ancestral tradition. A high fat diet using contemporary agricultural products is no more “natural” than HFCS.
The “paleo” part is just a marketing ploy, along the usual lines of “our ancestors had secret knowledge which we have forgotten [but I discovered it and will be sharing with you in this 300 page book which could be written in 20 pages but then people would not buy it at the airport].”
Evidence-based dietary guidelines are widely available and not that tricky or even expensive to follow (basically, limit sugars and booze, eat real food, especially veggies). The USDA (and similar agencies in every developed country) review and update them regularly, but they do not change substantially.
I do not understand why people go for fad diets, especially crazy ones. Eg the carnivore diet just looks like a recipe for colorectal cancer (meat, fat, no fiber).
I have difficulty believing that it is possible to consume 6-9lbs of cheese, butter, and ground meat per day. This would roughly be 19,000-29,000kcal per day. If this was possible, weight loss would only come through substantial diarrhea and no absorption of these calories. I’m skeptical of the details of this story.
Jd:
I was wondering that too. The journal article is very short and gave no details. I doubt it was 9 pounds per day. It was probably 9 pounds over some period of time that the patient recounted.
The paper is paywalled. The article and this blog both give the false impression – without ever saying it – that the man was eating a lot of cholesterol and that is why his serum cholesterol was high. The myth that animal cholesterol becomes human cholesterol never made any biological sense and has long since been debunked. In this case, I am guessing that his high fat diet exceeded the processing capability of his liver, but maybe someone who can get past the paywall can fill in the blanks.
Another thing. Is it even possible to eat nine pounds of cheese in one day? I can think of at least one bodily function that would shut down for me!
Matt:
I don’t know why you say that I “give the false impression – without ever saying it – that the man was eating a lot of cholesterol and that is why his serum cholesterol was high.” I mean, yeah, I never said it. Also I don’t see why you say I gave that impression. What I did say is that I think Seth was fooling himself, and others, when he thought he had good evidence that eating a stick of butter a day was improving his brain functioning.
Regarding the nine pounds of cheese: I was wondering that too. The journal article is very short and gave no details. I doubt it was 9 pounds per day. It was probably 9 pounds over some period of time that the patient recounted.
I can see behind the paywall. It’s not actually a regular article; it’s a “Cardiovascular Images” item, consisting of one paragraph, a figure, and a figure caption. There’s nothing referring to intake of cholesterol, but I don’t see any mention of cholesterol intake in the news article either. The implication in the item is that the high fat intake led to high blood lipid levels.
If you’re wondering what the images look like, just imagine tracing over all the wrinkles and folds in your hands and fingers with an orange-yellow highlighter.
This is only partially correct. Saturated fatty acids (at least most of them) seem to inhibit the LDL receptor in the liver, thus elevating serum lipid particle count – and therefore typically serum cholesterol as well, although it’s not the cholesterol per se that’s the problem, its the LDL particle count (best measured as serum ApoB protein). Small-modest amounts are probably fine for most people, and will be associated with dietary benefits when they’re coming along with nutritious whole foodstuffs. Mainlining them in a purified/concentrated form is a bad idea. Dietary cholesterol intake per se is a different beast. Roughly 1/3 of population are ‘hyper-absorbers’ for whom dietary intake does raise LDL/ApoB, while the other 2/3 are virtually immune to that effect, and therefore eating cholesterol per se won’t make much difference. So to sum up, dietary SFAs raise LDL/ApoB, especially if you mainline butter, coconut oil and so on, and dietary cholesterol has a variable impact depending on absorption. For the latter, Ezetimibe and dietary fiber can really lower serum LDL/ApoB.
Outside of extremes though, people generally over-estimate the impact of dietary pattern on serum lipids. If you really want to move the needle on LDL, check out PCSK-9 inhibitors like Repatha. I have a relative that went on Repatha, and reduced LDL from a high adult range down to a neo-natal level within 2 weeks.
I remember much of this history having seen it play out at the time. The self experimentation stuff was a clear positive step even if more skepticism was warranted.
On a related note, I’ve been thinking about starting my own diet fad. The basic idea is “avoid extremes”. For example, instead of trying to figure out which oils and fats are the best or worst using unreliable studies, just get a middling amount of oil/fat and rotate through all the different types.
The hope is that by simply moderating exposure to too much of any one type, while getting a normal amount overall, is a more reliable way to avoid problems than parsing research papers. Now to do some self experimenting …
You sure it was heart failure and not heart disease? Very different things. Heart failure is not nearly as strongly associated with cholesterol as heart disease.
The well-known phrase attributed to Eliezer Yudkowsky is “There only two kinds of diets: food and no food.”
Does it change the meaning to know that Yudkowsky is a bit portly?
Somewhat related, in that it involves what we consider “normal” food consumption, and it involves butter, I ran across on Facebook a post that showed the weekly meals of a family of English farm laborers in 1912. I dug up the source; link here: https://archive.org/details/howlabourerlives00rown/page/47/mode/1up
Tea, butter, and scarcity.
Ragu:
I went to the source that you listed
https://archive.org/details/howlabourerlives00rown/page/47/mode/1up
but, take a look: for four “dinners” per week, there were “meat, potatoes” and for two nights there was “stew” and one night meat soup with vegetables. The British term “dinner” seems to apply to a midday meal; the tea, butter (margarine) and scarcity appear to be the evening meal which in many societies, in contrast to present-day USA, is much lighter and perhaps much better for health.
With my last line I was making a quick quip, not a serious dietary assessment. In any case, the ingredient list, from a modern perspective, is remarkably minimal.
I vaguely recalled reading something about what the voyagers for the Hudson Bay Company ate, so I looked it up. “The voyageurs west of Grand Portage carried pemmican and supplemented it with wild rice (a kind of oat) when possible. A man doing hard labour required eight to ten pounds of fresh meat – which he had no time to hunt, dress and cook. Two pounds of pemmican nourished him as well. At the night camp he could hack off a chuck and boil it. For breakfast and the noon break he munched it dry. Most anticipated was the nightly tot of rum. ” https://calverley.ca/article/02-002-the-voyageurs-the-backbone-of-the-fur-trade/
Pemmican seems to be having a moment. Quite a few outdoor types are making it according to various homegrown recipes. I wouldn’t be surprised if it’s commercially available in some form. I’ve heard it’s pretty bad taste wise, even with berries to sweeten it.
Hard tack (ships biscuit) seems to be similarly resurgent and similarly bad tasking. Both would be fun to try once though.
When I was in ninth grade we did a school trip, hiking part of the Appalachian Trail and then canoeing part of the Potomac. Back in those days — 1979 — we didn’t have no fancy hiking food, nossir. But we did have commercially available pemmican. It was a chewy, grainy chunk about the size of a deck of cards (somewhat like a modern-day Clif bar), in a soft plastic-lined brown paper baggie with black or red writing. I think there might have been a picture of a stereotypical Indian head on the front (profile, feathered headdress) but I might be conflating the pemmican branding with something else. I don’t remember the brand name or anything like that. What I do remember is that wherever my parents took me to get me outfitted for the trip, pemmican was pretty much the only compact ready-to-eat food. You had to make your own GORP.
Anyway pemmican has been commercially available for a long time. But I wouldn’t be surprised if it is having a moment (again). Everything old shall be new again.
Thanks Phil, it looks like buffalo based pemmican is readily available, which is presumably closer to what the Indians ate.
I was thinking about making some using beef, but then I’d have to buy the beef and I’m not sure I want to flaunt my wealth in public like that.
A lot of years ago I was at a symposium on pre-clinical safety testing of biotech drugs that was held at the old Upjohn estate in Kalamazoo (Upjohn was still an independent pharma company at the time; the estate was magnificent with guest cottages and a stocked fishing pond where you could catch your own dinner if so inclined). I gave a talk on regulatory implications and it was at lunch on the final day when I had an interesting discussion with a preeminent toxicologist who was also a speaker. He was having desert and drinking coffee. I asked how he reconciled eating and drinking suspected carcinogens and his reply was “…eat lots of different things and in moderation…”; this is what my grandmother taught me. I’ve pretty much followed that advice and all my blood markers are in the good range (LOL).
Alan:
As discussed here, I’ve always been in good health but a few years ago my doctor informed me that my cholesterol level was persistently too high, and this scared me into changing my diet. Indeed, that set of events changed my whole take on Seth’s diet and similarly-touted miracle diet plans.
Looking at Seth Robert’s wiki page, there are some weird things that make me think that he really had no idea what properties an experiment needs to have (surprising for a psych professor). He just wildly jumps to conclusions:
> After many failures to see an improvement in his sleep, he eventually discovered that delaying breakfast, seeing faces in the morning, morning light, and standing solved this problem.
Seeing faces in the morning? The causal chain of events between seeing faces in the morning and being able to sleep at night must be pretty involved.
> In 2000, Roberts visited Paris. He noticed in himself a significant loss of appetite and speculated that it was due to experiencing unfamiliar flavors of soft drinks that were not available to him in the USA.
I wonder if it occurred to him that there could be many other equally (im)plausible explanations for loss of appetite in a foreign country.
I wonder why he didn’t put his conclusions to the test with controlled experiments. It’s perfectly fine to speculate and come up with theories based on hunches, but one has to take the next step before acting on them.
Shravan:
I agree, and I told Seth that. His take was as follows: (1) he’d tried lots and lots of things that hadn’t worked, therefore he was doubtful that he could’ve fooled himself or that his treatments could’ve worked through some sort of placebo effect; (2) he did have theories (based on set-points, paleo lifestyles, etc.) that explained his results; (3) it would be fine for people to do blinded experiments, but he felt that the results from his and others’ unblinded self-experiments were enough to convince him that his ideas worked, and he was more interested in learning new things.
I found Seth’s reasoning convincing enough that I basically believed that his treatments worked, at least for him and some others. I gradually lost my belief, due to several factors: (a) He started to make more and more outrageous claims based on weaker and weaker evidence (for example, the claim that eating a stick of butter a day improved his cognition, based on unblinded self-administered memory tests or something like that, which seems a lot more problematic than recording one’s sleep times and one’s weight); (b) the replication crisis in psychology made me realize how easy it was for researchers to fool themselves–oddly enough, one of the people who uncovered the replication crisis was Hal Pashler, a friend of Seth–I realized that if researchers could regularly come up with mistaken conclusions even in controlled studies, that this could be even more of a problem in uncontrolled experiments; (c) years later, when I decided to eat less and lose some weight, it made me understand how, if the decision to eat less were coupled with just about any magical-seeming treatment, it could make the treatment appear to be effective. I discussed that in my post, The rise and fall of Seth Roberts and the Shangri-La diet. Some commenters didn’t like what I had to say (see here and here)!
Wait, Roberts (and not Robert) is the same guy as this famous 2000 Roberts and Pashler paper, How persuasive is a good fit, in Psych Review? What the heck?
I make my students read this paper every single year; I think it’s the most important paper I ever read on evaluating model fit, and it had a huge impact on my life and my work. I just can’t believe that this is the same person who’s did all these nutty experiments and drew these nutty conclusions. Maybe Roberts should have gone back and read Roberts and Pashler 2000. That paper and Roberts’ thinking regarding thee things are not compatible.
“He started to make more and more outrageous claims based on weaker and weaker evidence ”
The most surprising thing about getting older is how often people go off the rails as they age. Everything from people’s politics drifting from reasonable into bizarre territory, to scientists latching onto weird views or numerology or whatever they never would have held when younger, to the incredible percentage of middle aged women in the US on anti-depressants, to extreme cynicism (“everything is a scam”), to extreme contrarianism (“everything experts say is a lie”), to conspiracy theorizing …
If you made it past fifty and you’re still normal, kudos to you.
In Seth’s case, I attribute some of this spiraling into extreme attitudes to the positive attention he was getting on the internet.
I did make it past 50 (in fact, I made it past 60 last year), and I do seem to have developed some of the characteristics you describe:
> extreme cynicism (“everything is a scam”), to extreme contrarianism (“everything experts say is a lie”)
Is this due to the ageing process and hardening of views, or is it because I spent 25 years studying a field in which a huge amount of scammy work got published, and I studied statistics and learned to recognize noise better? I don’t know. The contrarianism is the perennial disease of the academic, though. Although I have in fact noticed that my wife (married 34 years) now disagrees with literally anything I say, much more than before. (When I pointed this out to her, she told me I was wrong :).
What an outrageous and unsubstantiated claim! It is true that dementia strikes old people more than young ones, but that is a far cry from any generalization of age contributing to “going off the rails.” It is also true that older researchers commit more data falsification, just as they publish more papers, teach more students, and eat more bread – the older you live, the more of everything you accumulate.
Now, we’ve had discussions on this blog about the age of politicians and some people support an age limit (recognizing the advanced age of both Biden and Trump in the recent cycle). Now that we have Trump 2.0, and we find that Musk’s key engineers looking through government budgets are aged 19-24, I think it is reasonable to ask whether the dangers in politics are greater from the 65+ group or the under 25 group. I actually don’t have an opinion on the answer, but I think it is a valid question.
I dunno. I think the folks going off the rails bit looks way worse than it is because it tends to be the best physicists (and Nobel Prize winners) who go off the rails the most badly.
And at 72, I’m not into contrarianism or conspiracies. At all.
But the extreme cynicism bit is largely simply correct. Fusion energy* may be more like 30 years away instead of the 40 it’s been since I was in college, the idea that LLMs have something to do with “intelligence” is inanely stupid, “neural nets” bear no similarities whatsoever to actual neurons and their actual networks (but they are a great way of getting mileage (gradient descent in insanely high-dimensional spaces, especial when combined with domain representations (DeepFold)) out of SIMD computational devices (I just wish the smart people doing the good work would stop using the term “neural nets”, sigh.), remember “bubble memory” ??, ROFL. Blockchain. NFTs. The multiverse, multiple worlds, the anthropic principle (these are really really stupid (hint: Science is supposed to describe the world that we actually live in and have access to, not to write inane scifi), yet the physics hypesters blather on an on about them shamelessly). Etc. Etc. Etc.
I’ve even gotten a bit irritated with Science lately, because although they’re wonderful about checking in with folks not related to the work they’re reporting on, too many articles end with “this may lead to clinical treatments for X”, or “this may lead to commercial processes for extraction of lithium from brine”, when it’s clear that those are decades away. (I’m back from the 12 days in the hospital (playing the patient role in a Dr. House episode: 2 problems that interact. I told my lung doc about the silly send the workers to check out the patient’s house for toxins trope (used way too many times in Dr. House), and he said “We used to do that before Covid.” (!!!)) with a nasty version of pneumonia (how much is scar tissue and how much is clearable by prednisone will determine when I get the surgery for the other problem, and (while I’m in the hospital (!!)**) Science had an article on a newly discovered biomarker that determines whether you get scar tissue or new functioning lung tissue after lung injury, and that’s kewl, but it ain’t gona be clinically usefull in my lifetime, even though I seem to be negative for all the 10 biomarkers that indicate worsening disease…, so I’m being grumpy.)
In real life, real cases of actual depression are really really nasty (I’ve seen it: it’s truly horrific), but most “depression” is folks being rationally unhappy about the difficulty of their lives, and “antidpressants” have only been shown to be statistically (not clinically) effective in one, very specific, usage case (first diagnosis of one particuar type of depression), and in a lot of depression, simply walking is actually clinically effective with effect sizes far higher than anything ever seen with an antidepressant. Yet the last time I checked, over 11% of the US population was on antidepressants. IMHO this should be (and probably will be in time) seen as worse corporate mafeasance than opiods.
But being 72 means that the universe is playing russian roulette with my life: every year the probability of something random going wrong goes up. Like they say, the only thing worse than getting old is not getting old. But my SO has stopped complaining. I’m saying to myself “Hey, stuff happens, get over it”, and “Damn, the doctors are great, the hospital’s new and amazing, the nurses are great, the medicine’s way better than it was even 10 years ago, this is kinda fun.” Said SO is going “Sheesh, this is serious.”.
*: The LLL bit of getting more energy out of a laser-imploded pellet showed that there really is something there (and was great science), but almost no one reporting on it noted the enormous amount of energy required to and lost in creating the pulse that zapped the pellet, which itself was energetically costly to create. Sheesh.
**: Like a cat bringing a dead bloody rat or bird it had killed to show off to it’s owner, I love bringing docs articles from Science or whatever relevant to their work. Most are amused. My eye doc was so freaked by the contaminated OTC eye drops bit (some patients died, others blinded) that he missed my point that he had previously told me “Don’t even think about using OTC eye drops” and I was seriously thankful.
And I just noticed your comment about being “normal.” I have no idea what that is – if it is anything like “average,” then I thought we were beyond finding that a useful concept. When it comes to mental capacity and/or belief in misinformation, I don’t think “normal” is well-defined or useful. I think there is tremendous variability in people’s beliefs about strange or implausible things. Use of the term “normal” reveals more about the user of the term than anybody it might be applied to.
Shravan –
(When I pointed this out to her, she told me I was wrong :).
I remember being wrong once, but I could be wrong about that.*
*(apparently from the Monkey Wrench Gang)
Talking to myself but, the absoulute insane inanity of the LLM idea is on full dress-parade display.
https://garymarcus.substack.com/p/shame-on-google-twice
(Andrew: sorry to be ranting on a post about your friend Seth. My friend Alan, who taught me a lot about assembler language programming when I was an undergrad was bipolar, and I remember being hyper-enthused about something and runing up the stairs between the 8th and 9th floors at Tech Square, and Alan was coming down the stairs. ONE.STEP.AT.A.VERY.SLOW.TIME. I could tell by his eyes that he recognized my excitement and realized he couildn’t respond. It was the most horrifically sad expression I’ve ever seen on a human face. Friends are valuable, no matter how nasty their problems.)
Andrew, don’t worry, it’s not your fault. It’s Trump’s fault.
Gb:
Please save such empty comments for twitter, 4chan, and your local newspaper’s comment threads. Thank you for understanding.