“Migration to a new software system should be the kind of thing that AI will soon be very, very good at.”

Quite the opposite IMO. Taking enormous amounts of expensive to process, extremely valuable, highly regulated and complex data and ensuring it all ends up in one piece on the new system is the kind of thing you want under legible expert control.

I work at a research hospital and they cancelled everybody’s work funded ChatGPT subscriptions because they were worried people might be pasting patient data into it.

Why despair about refactoring economic regulations? Has every angle been exhausted? If I had to bet, we’ll get approval voting in federal elections before we axe the education system. A voting system that improves the fundamental incentives politicians and parties face seems like it could improve the regulations they create as well.

Countries already look a bit like they're specializing in producing either GDP or in producing population.

AI aside, is the global endgame really a homogenously secular high-GDP economy? Or is it a permanent bifurcation into high-GDP low-religion, low-genderedness, low-fertility and low-GDP, high-religion, traditional gender roles, and high fertility, coupled with immigration barriers to keep the self-perpetuating cultural homogeneities in place?

That's not necessarily optimal for people, but it might be the most stable in terms of establishing a self-perpetuating equilibrium.

Is this just an extension of partisan sorting on a global scale?

Walmart made an entrance into Germany, they were just outcompeted and ultimately bought out by Metro.

https://learn.saylor.org/mod/page/view.php?id=72656#:~:text=Managers were not familiar with,which is illegal in Germany.

Some small experiments related to this effect. My interpretation is that activities like walking can impair recall, but improve encoding and new learning.

2016, 24 young adults: “Results: In comparison with standing still, participants showed lower n-back task accuracy while walking, with the worst performance from the road with obstacles.”

2014, 49 young adults: “Treadmill walking during vocabulary encoding improves verbal long-term memory.”

2014, 20 young adults: No significant difference in a spatial working memory task for any walk speed, including standing still.

2021, 11 people with MS-related impairments in new learning: Moderate to large improvement in a verbal learning task.

2011, 80 college students: “ Walking before study enhances free recall but not judgement-of-learning magnitude.”

1. https://www.frontiersin.org/journals/behavioral-neuroscience/articles/10.3389/fnbeh.2016.00092/full
2. https://link.springer.com/article/10.1186/1744-9081-10-24
3. https://www.frontiersin.org/journals/human-neuroscience/articles/10.3389/fnhum.2014.00288/full
4. https://www.sciencedirect.com/science/article/pii/S1551714421002998?casa_token=HrgrESH2FzEAAAAA:TVXtI20lXKux0wnOUGlM_iONup8gslQ1lsGd8dxa0QlWAhN1XFA-pGK6xWxYYQkkYJgca2MnLg
5. https://www.tandfonline.com/doi/abs/10.1080/20445911.2011.532207

Tracing Woodgrains' tweet reveals Johnson to be brutal and profoundly manipulative. Why think he only acts that way toward his wife, not his customers? Why be curious about the health advice offered by a person like that?

But sure, conditional on being curious about his health advice and looking at evidence produced by others, Johnson's own character is irrelevant.

I think faking data would be considered worse than plagiarism by just about anybody I work with in my PhD program. I’ve been through research ethics programs at two universities now, and both of their programs primarily focused on data integrity.

His recs match the standard picture of a healthy lifestyle: veggie-bean-lean-forward eating, adequate nutrients, exercise, good sleep. Following his recommendations seems fine? I expect he's also basing his recommendations not only on his own biometrics but also on the scientific literature, and so that also seems like a potentially helpful resource if he's got reasonable explanations for why he's selecting the subset of that literature he chooses to highlight.

Evidence his system can motivate and provide superior results to other diet-and-exercise regimens on the basis of his own personal results is, of course, massively confounded.

He's selling the supplements he recommends, he's extremely rich, he's unmarried (though has 3 kids, I don't know his involvement), he's being danced around by doctors all the time as far as I can tell, I expect he's outsourcing a lot of his domestic labor, and he has chosen a line of work where he's professionally invested in a low-stress, healthy lifestyle. He's clearly conscientious and extremely smart given his prior success in business. He probably wouldn't have blown up on the internet if he didn't happen to look young and fit. I question whether exposure to his protocols is any better at causing behavior change for the better than alternative systems, and there are intense selection effects for who chooses to and succeeds at following his protocol (and it's not just selection for the "disciplined and capable"). 

These are the fundamental challenges with trying to interpret n=1 longitudinal data. It's hard to update on unless you're a lot like the test subject. And this test subject is factually weird, so you're probably not like him. That doesn't make is ideas bad, it makes his evidence almost worthless to nearly everybody except him.

The reason his recs make sense is because they're drawing on a tremendous amount of standard scientific research. That information, in principle at least, you already had access to without him. So his n=1 longitudinal data seems more like a driver of the narrative and excitement around his brand than a meaningful point of evidence in favor of his specific lifestyle plan.

I think the answer is simply that the modern world allows people to live with poverty rather than dying from it. It’s directly analogous to, possibly caused by, the larger increase in lifespan over healthspan and consequent failure of medicine to eliminate sickness. We have a lot of sick people who’d be dead if it weren’t for modern medicine.

Fungal infections are clearly associated with cancer. There's some research into its possible carcinogenic role in at least some cancers. There's a strong consensus that certain viruses can, but usually don't, cause cancer. Personally, it seems like a perfectly reasonable hypothesis that fungal infections can play an interactive causal role in driving some cancers.  In general, the consensus is you typically need at least two breakdowns of the numerous mechanisms that regulate the cell cycle and cell death for cancer to occur.

I'm a PhD student in the cancer space, focusing on epigenetics and cancer. Basically, this is the field where we try to explain both normal cellular diversity (where DNA mutations are definitely not the cause except in very specialized contexts like V(D)J recombination) and cancers apparently not driven by somatic mutations in protein-coding genes.

Mutations not in protein-coding genes are not necessarily inert. RNA can be biologically active. Noncoding DNA serves as docking sites for proteins, which can then go on to affect transcription of genes into mRNA. The proteome can also be affected by alternative splicing of mRNA. Non-coding mutations can potentially affect any of these processes and thereby affect the RNA and protein landscape within a cell.

In 2024, our ability to detect mutations varies widely across the genome, due both to the way we obtain sequencing data in the first place and the way we attempt to make sense of it. NGS sequencing involves breaking the genome into short fragments and reading around 150 base pairs on either end of the fragments, then trying to map it back to a reference genome. Mapping quality will suffer or completely degrade both if the patient has substantial genetic difference from the reference genome or in regions that are highly repetitive within the genome, such as centromeres. When I work with genetic data, there are regions spanning multiple megabasis that are completely blank, and a large percentage of our reads have to be thrown out because we can't unambiguously map them to a particular location on the genome. This will be partially overcome in the future as we start to use more long-read sequencing, but this technology is still in its early stages and I'm not sure it will completely replace NGS for the foreseeable future.

In the epigenetics space, we focus on several aspects of cell biochemistry apart from DNA mutations. The classic example is DNA methylation, which is a methyl group (basically a carbon atom) present on about 60% of cytosines (C) that are immediately followed by guanine (G). The CpG dinucleotide is heavily underrepresented relative to what you'd expect by chance, and its heavily clustered in gene promoters. Methylated CpG islands in promoters are associated with "off genes". The methylation mark is preserved across mitosis. It's thought to be a key mechanism by which cell differentiation is controlled. We also study things like chromatin accessibility (whether DNA is tightly packaged up and relatively inaccessible to protein interactions or loose and open) and chromatin conformation (the 3D structure of DNA, which can control things like subregion localization into a particular biochemical gradient or adjacency of protein-docking DNA regions to gene promoters).

These epigenetic alterations are also thought to be potentially oncogenic. Epigenetic alterations could potentially occur entirely due to random events localized to the cell in which the alterations occur, or could be influenced by intercellular signaling, physical forces, or, yes, infection. If fungal infections control cells like puppets and somehow cause cancer, my guess is that it would be through some sort of epigenetic mechanism (I don't know if there are any known fungi that can transmit their DNA to human cells).

Epigenetics research is mainstream, but the technology and data analysis is comparatively immature. One of the reasons it's not more common is that it's much harder to gather data on and interpret than it is to study DNA mutations. Most of our epigenetics methods involve sequencing DNA that has undergone some extra-fancy processing of one kind or another, so it's bound to be strictly more expensive and difficult to execute than plain ol' DNA sequencing alone. Compounding this, the epigenetic effects we're interested in are typically different from cell to cell, meaning that not only do you have these extra-challenging assays, you also need to aim for single-cell resolution, which is also either extremely expensive (like $30/cell, isolating individual nuclei using a cell sorter and running reactions on each individually, leading to assays that can cost millions of dollars to produce) or difficult (like using a hyperactive transposase to insert DNA barcodes into intact nuclei that give a cell-specific label the genetic fragments originating from each cell, bringing assay costs down to a mere $50,000-$100,000 driven mainly by DNA sequencing rather than cell processing costs). This data is then very sparse (because there's a finite amount of genetic information in each cell), very large, and very difficult to interpret. We also have extremely limited technologies to cause specific epigenetic changes, whereas we have a wide variety of tools for precisely editing DNA.

For potentially oncogenic infections, fungal or otherwise, you'd want to show things like:

• We can give organisms cancer by transferring the pathogen to them
• We can slow or prevent cancer by suppressing the putatively oncogenic pathogen.
• The pathogen is found in cancer biosamples at an elevated rate
• There are differences between the cancer-associated pathogens and non-cancer-associated pathogens, or cellular changes that make them more susceptible to oncogenesis through their interactions with the pathogen

All of this seems like a perfectly respectable research project, just difficult. I can't imagine anybody I work with having a problem with it. Where they probably would have a problem would be if the argument was that "fungal infections are the sole cause of cancer, and DNA mutations or epigenetic alterations are completely irrelevant to oncogenesis."

There's an angle I've neglected in this post until now, which is the perspective from evolutionary theory. it's more common to refer to this in explaining how cancer evolves within an individual. But it's also relevant to consider how it bears on the Peto paradox. Loosely, species tend to evolve such that causes of reproductive unfitness (including death) tend to balance out in terms of when they occur in the life cycle. Imagine a species under evolutionary pressure to grow larger, perhaps because it will allow it to escape predation or access a new food source. If the larger number of cells put it at increased risk of cancer, then at some point there would be an equilibrium where the benefit of increased size was cancelled by the cost of increased oncogenesis risk. This also increases adaptive pressure to stabilize new oncopreventative mechanisms in the population that weren't present before. This may facilitate additional growth to a new equilibrium.

This helps explain why cancer isn't associated with larger size: adaptive pressure to develop new oncopreventative mechanisms increases in proportion to the risk to reproductive fitness posed by cancer.