Apple's Liquid Glass is prep work for AR interfaces, not just a design refresh (omc345.substack.com)

The problem with any Alzheimer's research is that the disease is still not well understood. It is likely that multiple diseases are being bundled in as a single disease. The tests for AD, are somewhat rudimentary. Beginning with psychological and neurological tests, the blood work to rule out other conditions, followed by a PET scan to look for brain atrophy, and CSF measures for amyloid and tau levels.

It seems almost like they're basically ruling out any disease we can actually measure for and then if it isn't one of those, it's AD.

Does this mean the Amyloid hypothesis is wrong? Unlikely. Is it incomplete? Absolutely!

But articles shouting that all the research should be thrown out are not helpful .

The AD community know that they don't understand the disease, and though therapeutics have been mostly focused on amyloid and tau, research into how the disease works continues.

andrewflnr · 10h ago

> But articles shouting that all the research should be thrown out are not helpful .

Good thing OP isn't one, then. In fact it makes a pretty similar point: all the non-amyloid research also should not have been thrown out. Or rather, killed before it got that far; you can decide whether that's equivalent or worse.

Nopoint2 · 9h ago

It seems involves some kind of metabolic disturbance, is there any explanation for this? https://academic.oup.com/metallomics/article/9/8/1106/601364...

cogman10 · 13h ago

Could the Amyloid/tau hypothesis be a cart before the horse situation? It is my understanding that the current hypothesis is that the buildup of these proteins causes Alzheimer's. Could it be that Alzheimer's causes these proteins to build up?

daveguy · 12h ago

It could be. That's why we need more research. Well funded and not vilified.

PaulKeeble · 14h ago

Its not the only disease this has happened to either. ME/CFS has been railroaded by European governments that funded only psychological research despite numerous pathological findings and better theories of the disease, this prejudiced treatment started in the 1970s and persists to this day including the corruption of the PACE trial results which researchers tried to hide the data of.

Research fraud in medicine is alarmingly common.

ahazred8ta · 9h ago

For a long time, fibromyalgia was only diagnosed by ruling out everything else. There was a lot of 'It's all in your head.' Last year they developed an actual blood test for it that detects the wonky immune system response.

incangold · 13h ago

One infuriating thing about PACE is that even the fraudulent results only showed a 22% recovery rate.

For a disease as serious as ME/CFS, a treatment with a 22% recovery rate is far from good enough. Even if PACE stood up to scrutiny it wouldn’t have made sense to give up on finding better treatments.

noitpmeder · 11h ago

I mean, 22% sounds pretty damn good if there are no long lasting negative side effects of being part of the remaining 78%.

Like, sure, shoot for 200% cure rate, but even a success rate of 1% cured of a previously unrecoverable situation is insanely informative.

noitpmeder · 11h ago

(self reply) (obviously if the 22% number itself is bogus then you can't trust anything)

Henchman21 · 14h ago

There’s money to be made; don’t let facts stand in the way of profits!

Centigonal · 14h ago

This argument gets invoked a lot when it comes to medical dishonesty, but I really don't think it applies in the case of ME/CFS. If we could find the pathology behind the condition, there is huge money to be made in pharmaceutical treatments. Just look at the enormous amount of money being made treating auto-immune illnesses with Humira/Skyrizi/Xeljanz/etc, treating diabetes with GLP-1 agonists and CGMs, and treating obesity with GLP-1 agonists (and depression before all that!). Sometimes treating the chronically ill is the most profitable option.

cjbgkagh · 11h ago

ME/CFS shows up as immune system dysfunction and for most people is likely downstream from dysautonomia. Given the huge overlap of people with long covid and hEDS, and that long covid presents as dysautonomia, the math likely works out that most people with ME/CFS actually have hEDS. It’s massively underdiagnosed. It’s likely one of many TNXB SNPs, these are overlooked for a few main reasons - it’s considered too common for a rare condition and it’s hard to sequence and only detectable with modern high quality deep WGS.

The other thing is that GLP-1As actually do appear to help with autoimmune conditions and has been, in my case and in the case of many others, one of the most effective medications for treating ME/CFS.

I wish it was a lack of a profit motive that caused medical researchers to be so off base but it appears what is more likely that they’re not very good with stats, as being good at stats would get in the way of all the ‘accidental’ p-hacking they’re so addicted to.

Edit: Added an email to my hacker news user details

mjevans · 10h ago

Most of us (myself included) are not medical doctors specialized in this field. Could you please expand your acronyms? I only know of CFS (if it's the same one) because a cousin has it.

cjbgkagh · 9h ago

Sure,

  ME/CFS: Myalgic Encephalomyelitis / Chronic Fatigue Syndrome
  hEDS: Hypermobile Ehlers Danlos syndrome
  TNXB: A gene for making the tenascin-X protein, it's one of the RCCX (Receptor of Complement component 4) gene cluster genes.
  SNPs: Single nucleotide polymorphisms
  WGS: Whole Genome Sequencing

ahazred8ta · 9h ago

hEDS, hypermobile Ehlers-Danlos syndrome, a joint problem with co-occurring immune system problems, sometimes caused by defects in the TNXB gene https://duckduckgo.com/?q=hEDS+hypermobile+Ehlers-Danlos+syn...

Myalgic encephalomyelitis/chronic fatigue syndrome https://en.wikipedia.org/wiki/Myalgic_encephalomyelitis/chro...

s5300 · 11h ago

Could you drop a form of contact? I have some questions about your knowledge of MECFS (especially regarding your comment about test cyp. & hgh for CCI)

gleenn · 13h ago

I agree, but will the money be made by the same people? Like, if a psychologists group stands to benefit from treating it they wouldn't reap the benefit of the pharma research. Kind of like how the alcohol industry lobbies against weed or other possible substitutes. Or the car companies buying out the rail roads and then shutting them down to prevent competition. There is just so much inefficiency from competing industries sabotaging each other. You are also making a leap that understanding the cause is physiological means that there is a drug you could sell and not cost billions of dollars to get to market without risk. I always wonder about the PReP drugs for treating AIDS was really the best outcome instead of aiming for the cure. Definitely tradeoffs with a lot of money one way or another.

nradov · 13h ago

The psychologist lobby is not as powerful as you seem to think. Pharmaceutical companies have brought numerous drugs to market that have proven to be more effective than talk therapy for certain serious chronic mental health conditions.

Centigonal · 13h ago

I think that's reasonable. I think there's definitely an economy-wide bias by vested interests against disruption, which gets us stuck in local maxima, sometimes for centuries.

seltzered_ · 14h ago

- The great brain clearance and dementia debate - https://www.nature.com/articles/d41586-025-00962-y (2025)

- The Devastating Legacy of Lies in Alzheimer’s Science - https://www.nytimes.com/2025/01/24/opinion/alzheimers-fraud-... (2025), referencing the book Doctored https://en.m.wikipedia.org/wiki/Doctored_(book)

- The maddening saga of how an Alzheimer’s ‘cabal’ thwarted progress toward a cure for decades - https://www.statnews.com/2019/06/25/alzheimers-cabal-thwarte... (2019, by the late Sharon Begley)

tsoukase · 2h ago

Every degenerative CNS disease is caused by death of non replicating cells. By definition, the cure of such diseases is equivalent to immorality. This already highlights the importance and difficulty of such achievement.

The amyloid hypothesis seems too simple and superficial to account for a decades long process (it is speculated that Alzheimer's starts up to 20 years before symptom onset). The ultimate problem is to find the underlining cause(s) and not correlations.

djoldman · 14h ago

Derek Lowe's take:

https://www.science.org/content/blog-post/reaction-alzheimer...

throwawaymaths · 14h ago

he was not in the trenches in 2003. in 2003 we were working in an Alzheimer's lab and everyone in our lab at least was expressing suspicion that there was something wrong with the hypothesis. pretty much every internal lab meeting started with "the amyloid hypothesis is... [this statement exists because our funding stream] though it is not conclusively proven, wink wink"

caycep · 13h ago

Granted, I have heard a scientist (in an unrelated field, I think it was some astronomical NPR interview) describe science as a bit like a supertanker...there usually is some prevailing thing that everyone believes in, but as contrary evidence piles up, the direction slowly turns.

I guess my own question is whether Alzheimer's/amyloid thinking was atypically stuck on one hypothesis, vs. is this just the slow pace of progress as usual for a given field? I mean...it's not like the amyloid deposition isn't there...

I only play a AD expert on TV (haha I jest...I like to say this because I had no intention of specializing in this when I was training but, hey, in the real world, you have to treat the "market" that rolls in the door....). I work more in the Parkinson's world, and while I would say there are cliques, which do affect who gets NIH (or used to get...I have no idea what's going on there now...), I can't say there's one prevailing "cabal" that's obsessed with any one direction. the bigger issue is that current Parkinson's research is a bit scattershot in too many directions.

My other pet peeve is somewhat unrelated, where the article mentions other directions like neuroinflammation and oxidative stress; the problem is these are also vague/broad topics, that have been thrown around like panaceas for every disease from head to toe; my own superstition is that when a new drug candidate comes out for "neuroinflammation" or "oxidative stress", I'd bet a healthy bunch of nickels it won't amount to much.

nick__m · 13h ago

Do you see any curative Parkison (not even reversing the damage but stopping it) treatment being available in the next few years ?

jcranmer · 13h ago

His blog has a category just for discussing Alzheimer's, and he's been talking about ever since he started blogging. So here's a post from 2002 where he points out (somewhat obliquely) that amyloid isn't a proven hypothesis: https://www.science.org/content/blog-post/alzheimer-s-vaccin...

> This is looking like one of the crazy ideas that just might work - stipulating, for the moment, that amyloid really is the cause of Alzheimer's. . .

wileydragonfly · 14h ago

Which is the only good take… if there wasn’t “something promising” with the idea, people wouldn’t have kept at it. There’s only so much time you’re going to waste on pure fantasy before you move on. The whole ordeal had me quite upset as I was supporting studies looking at this, and the patients are beyond desperate for any treatment. None of us got rich.

robwwilliams · 14h ago

Great and highly critical book on this general topic: “How Not to Study a Disease” by Karl Herrup from MIT Press:

https://direct.mit.edu/books/book/5216/How-Not-to-Study-a-Di...

The core problem is much older than stated in this focused review.

fnordpiglet · 14h ago

This is from 2022

searine · 11h ago

Anti-science types keep holding this up as some kind of 'gotcha' or as a waste, but in the end it shows that the scientific method and the scientific establishment work efficiently.

Despite even intentional fabrication, the truth of it was found in a few years and the field marches on.

Cornbilly · 7h ago

HN (and Silicon Valley) has a contingent of people that want to attack the credibility of the scientific community so that they can present their own (usually very flawed) conclusions as being legitimate.

gertop · 7h ago

> it was found in a few years and the field marches on

I agree but 16 years is still significant. It represents 5% of the modern medicine era.

m3kw9 · 12h ago

So why do researchers is still hell bent on believing the beta amaloyd theory? I get voted down every time I ask this

xkcd-sucks · 11h ago

There's a lot of researchers in a lot of roles and it turns out finding targets is just one of those roles, the rest focus on optimizing therapies against those targets and they like having well defined targets to work on.

abetancort · 14h ago

Old and outdated.

esafak · 14h ago

How so?

trhway · 13h ago

Something is very wrong and corrupt if a large field of science, an army of scientists spending billions of dollars, is ruled by one easily repeatable study/experiment, and yet nobody cares to repeat it.

mousethatroared · 12h ago

Scientists do repeat experiments, they just don't publish the results because no one cares.

Specifically, results are replicated wherever you want to build on something someone has done. For example I'm making specific glass off of a Nature paper for a totally unrelated use.

If the glass doesn't work out will I publish my results? No, too much work to get rejected.

trhway · 11h ago

>Specifically, results are replicated wherever you want to build on something someone has done.

yes, you want to test your Alzheimer treatment, and you measure the amyloid in the patients before the treatment, and you don't find it in some patients... Are you not going to publish that?

Nasrudith · 12h ago

Nobody ever got tenure for repeating an existing experiment and getting the same results. It is a problem of incentives. Aside from doing something expensive like grant terms including potentially 'drafting' some scientists to attempt to replicate reproducible studies for funding as their next project I'm not sure what could be done to fix it.

porridgeraisin · 6h ago

Experiments are repeated if someone else is looking to expand on the work actually... it's just that if it doesn't work, no one wants to go to the effort of publishing that since it's a lot of effort and there'll be politics with the original author and there's the possibility that it was just a "push it through to satisfy the metrics" paper that every researcher "allows" every other researcher purely out of sympathy and there's the review process and deadlines and rebuttals and more deadlines and...

I suppose if you only value this entire end to end process then the experiment is not being repeated at all, but yeah, most things just get discarded when they don't work.

macinjosh · 13h ago

gotta trust that science yo

readthenotes1 · 12h ago

No no no.

Trust the Scientist! For they are pure of heart and unswayed by pride, greed, lust, avarice, or envy

richarlidad · 12h ago

Cash rules everything around me.

Model Once, Represent Everywhere: UDA (Unified Data Architecture) at Netflix (netflixtechblog.com)

Unsupervised Elicitation of Language Models (arxiv.org)

Peano arithmetic is enough, because Peano arithmetic encodes computation (math.stackexchange.com)

Last fifty years of integer linear programming: Recent practical advances (inria.hal.science)

SIMD-friendly algorithms for substring searching (0x80.pl)

The Many Sides of Erik Satie (thereader.mitpress.mit.edu)

Cray versus Raspberry Pi (aardvark.co.nz)

Solidroad (YC W25) Is Hiring (solidroad.com)

Slowing the flow of core-dump-related CVEs (lwn.net)

Endometriosis is an interesting disease (owlposting.com)

Filedb: Disk-based key-value store inspired by Bitcask (github.com)

Saab achieves AI milestone with Gripen E (saab.com)

TimeGuessr (timeguessr.com)

Implementing Logic Programming (btmc.substack.com)

Me an' Algernon – grappling with (temporary) cognitive decline (tidyfirst.substack.com)

Self-Adapting Language Models (arxiv.org)

How I uncovered a potential ancient Rome wine scam (phys.org)

Liquid Glass – WWDC25 [video] (developer.apple.com)

Student discovers fungus predicted by Albert Hoffman (wvutoday.wvu.edu)

The Army’s Newest Recruits: Tech Execs From Meta, OpenAI and More (wsj.com)

The international standard for identifying postal items (akpain.net)

Protecting your code from other people's bugs (doi.org)

If the moon were only 1 pixel: A tediously accurate solar system model (2014) (joshworth.com)

I convinced HP's board to buy Palm and watched them kill it (philmckinney.substack.com)

Whatever Happened to Sandboxfs? (blogsystem5.substack.com)

The Hat, the Spectre and SAT Solvers (2024) (nhatcher.com)

Mollusk shell assemblages as a tool for identifying unaltered seagrass beds (int-res.com)

Apple's Liquid Glass is prep work for AR interfaces, not just a design refresh (omc345.substack.com)

100 years of Zermelo's axiom of choice: What was the problem with it? (2006) (research.mietek.io)

Google Cloud Incident Report – 2025-06-13 (status.cloud.google.com)

Frequent reauth doesn't make you more secure (tailscale.com)

Jemalloc Postmortem (jasone.github.io)

$100 Hamburger (en.wikipedia.org)

When random people give money to random other people (2017) (quomodocumque.wordpress.com)

Green Tea Garbage Collector (github.com)

Show HN: Tattoy – a text-based terminal compositor (tattoy.sh)

Using computers more freely and safely (2023) (akkartik.name)

OxCaml - a set of extensions to the OCaml programming language. (oxcaml.org)

High-speed fluorescence light field tomography of whole freely moving organisms (opg.optica.org)

Caltrain official lived in secret apartment built illegally inside train station (sfstandard.com)

Ask HN: How do I give back to people helped me when I was young and had nothing?

AI agent startups at Y Combinator’s Spring ’25 Demo Day (businessinsider.com)

Shaping Light – Volumetric Lighting (blog.maximeheckel.com)

Paleoproteomic profiling recovers diverse proteins from 200yo human brains (phys.org)

How I program with agents (crawshaw.io)

GPU-accelerated Llama3.java inference in pure Java using TornadoVM (github.com)

Geometry from Quantum Temporal Correlations (arxiv.org)

A Study of the Winston Red: The Smithsonian's New Fancy Red Diamond (gia.edu)

EDAN: Towards Understanding Memory Parallelism and Latency Sensitivity in HPC [pdf] (spcl.inf.ethz.ch)

Meta invests $14.3B in Scale AI to kick-start superintelligence lab (nytimes.com)

How the Alzheimer's Research Scandal Set Back Treatment 16 Years (2022)

Comments (45)