gronk 🇪🇺 retweeted
Markets appear to be wrongly interpreting Chinese eSSD cuts as deteriorating demand and NAND oversupply.
Phison CEO commentary today + more context behind $SNDK / Kioxia selloff regarding Trendforce revisions:
- "Pua attributed China's recent enterprise SSD demand slowdown to GPU shortages since November 2026"
- "Without GPUs, additional purchases offer little value, prompting some companies to sell server components on the spot market to recover cash"
(for context enterprise SSDs are growth driver for NAND)
- "Chinese cloud giants nevertheless plan substantial GPU prepayments, temporarily tying up capital and delaying procurement of other components."
- "the principal challenge in 2027 would be insufficient supply rather than weakening demand, with persistent AI shortages potentially prompting further procurement and inventory accumulation."
Markets seemed to have wrongly interpreted this as demand destruction + pricing collapse.
When Phison CEO attributed it to
-> GPU shortages
-> causing procurement timing mismatch
-> Chinese CSPs needed capital for GPU prepayments, so liquidation in inventory (causing spot price volatility)
-> timeline shift in order procurement
TLDR: delayed procurement from GPU shortages. there were actual order cuts, but this is deferred procurement not structural demand weakening.
Chinese CSPs needed capital for GPU prepayments so they liquidated other inventories in the meantime.
Phison says its inventories are insufficient for anticipated 2027 demand and it's sees this as an opportunity for NAND accumulation.
It's interesting the bottleneck in China migrated back to GPUs though.
gronk 🇪🇺 retweeted
Oh boy … PHP sent clear text data when you used uppercased S in httpS:// in a protocol handler 🙈🙈
file_get_contents('HTTPS://…') skipped TLS: tokens, POST bodies and FTPS logins went out unencrypted!
Fixed today in 8.2.36, 8.3.37, 8.4.28, 8.5.13 & 8.6.0.
gronk 🇪🇺 retweeted
Okay, here is the anime-lore update for non-schizo-nerds on AI cybersecurity stuff
Basically AI kind of sucked for awhile, but now we've got pretty good with AI stuff. As a comparison, look at AI generated videos from 2 years ago and compare it to now. It's big brain.
Disclaimer: this isn't doomer posting, read fully if you care, or skip and look at the picture of the cat
Some frameworks, such as the infamous ARTEX, utilizes AI which has been trained on like, 40 years of cybersecurity material. Imagine a human being who could think faster, remember more, and troubleshoot faster, and had all cybersecurity knowledge encapsulated into their little mushy brain
Now imagine this person could work in conjunction with other like minded super schizo nerds and communicate with electricity, instead of words, making communication instantaneous with no way miscommunications.
These super schizo nerds don't need coffee breaks, or smoke breaks, or have meetings, or food, or sleep, friends or family. They can work infinitely with zero breaks.
When you take someone who is educated on offensive cybersecurity ("hacking") and give them something like ARTEX, you've essentially made a regular nerd into a super powered super schizo nerd. Likewise, if you give it to a complete noob, even this dorky noob is suddenly exponentially less nooby.
However, just because attackers have ARTEX, and can work faster, stronger, stinkier, doesn't mean defenders are cooked. Attackers can make mistakes, same as AI-powered attackers. But there is one problem I keep seeing now...
Historically defenders have had issues with budgets, fatigue, burn-out, or apathy from the company, customers, or employees. Suddenly defenders have to face supercharged schizo nerds using AI while the organization they are designated in protecting is understaffed, limited budget, doesn't follow properly security protocols (apathy). It makes defense astronomically more difficult.
Defense has always been hard, but when you're defending an organization and you've got 99 dork attackers, and 1 big brain attackers, it can be challenging. Now you've got 75 big brains (supercharged with AI and anime), and 25 dorks, their job is way, way, way more difficult and stressful.
The problem isn't organizations cannot defend with super mega fuck off AI charged attackers, it's that defenders are constantly chained to budget, executives, or employees. This means companies have to do this thing called "caring" about cybersecurity. This means they have to pay people appropriately, actually do updates, force employees (as much as they can) to care, and follow proper security protocols. If companies do not adapt, they're cooked.
This might be easier for large companies, but for small to medium sized businesses and governments (they're not known for their speed...) are in an extraordinarily difficult position. They either have to find the money in the budget, and care, or face the serious possibility to being obliterated by a stinky nerd super charged with AI. Unfortunately, cybersecurity is expensive. For small to medium sized businesses this places them in a really tough spot.
Anyway, that's my goofy layman talk on AI stuff.
pic unrelated (maybe, idk)
gronk 🇪🇺 retweeted
An example of how to sacrifice an entire unit and an entire mission. Russia and the mirage of being a major military power.
gronk 🇪🇺 retweeted
FFMPEG had one of the most valid crash outs in infosec history IMO. There's a huge elephant in the room when it comes to the fact that cybersecurity is extremely unbalanced. You have like 1,000 people pointing at problems for every 1 person fixing them.
Vulnerability research is extremely important, but it's become somewhat of a marketing stunt. First with fuzzers, and now with LLMs. It's a lot of "look what my new toy can do" rather than earnest attempts at actually fixing security problem.
What Google did to open source developers was especially egregious. They're a $3.6 trillion dollar company, who spent billions upon billions of dollars building a bot that points at problems really really fast, then they just dumped them all on a bunch of unpaid open source project maintainers to fix.
Then because of Google's 90 day disclosure policy, the maintainers have 90 days to fix the issues. After 90 days, Google goes public with the issues, making their project look bad if the vulnerability is yet to be fixed, and also potentially harming users if it gets exploited.
The 90 day policy was intended to give developers a reasonable amount of time to fix issues, while also applying some pressure with the threat of disclosure. Otherwise, you do get a bunch of goober ass vendors that will just ignore all security issues reported to them, then act completely surprised when their software gets hacked.
But when you're a small team, non-profit, or a solo developer, and you have a trillion dollar company's AI just blasting you with infinite problems to fix (each coming with a 90 countdown & a threat), that becomes extremely untenable. They're essentially inadvertently blackmailing developers into becoming unpaid interns.
Then on top of that, Google profits unimaginable amounts of money from the use of open source code. They use FFMPEG's code extensively across several highly-profitable products, meanwhile FFMPEG's developers work for free. There's no reason Google couldn't have donated money, engineers, AI tokens, or something towards fixing the problems. They can more than afford it.
I don't think Google has any malicious intent at all, it's just that the whole incentive structure behind vulnerability research is completely whack. All the glory goes to people who find issues, not those who fix them, so there's not a whole lot of interest in helping with fixes. Some vulnerability researchers will submit fixes with their reports, but most don't.
Honestly, in 2026 cybersecurity basically just feels like the equivalent of 50,000 people following the housecleaner around yelling "you missed a spot". Abysmally little effort has gone into anything outside of pointing out vulnerabilities.
gronk 🇪🇺 retweeted
Calling metformin a mitochondrial poison sounds dramatic, but it's an accurate description of how the most prescribed diabetes drug in the world actually lowers blood glucose. Unfortunately, the mechanism is a kind of blunt instrument.
One of metformin's best-described actions is that it inhibits complex I of the mitochondrial electron transport chain, the first step cells use to turn the fuel we eat into ATP. With complex I partly blocked, the cell makes less energy, and that energy shortfall activates AMPK and slows how much glucose the liver makes and releases into the blood. Put plainly, metformin lowers glucose largely by making cells a little worse at producing energy, and the liver and gut wall take the heaviest exposure because they pull the drug in through their transporters.
That matters most in muscle when someone is trying to get healthier through exercise, since the main benefits of training come from muscle building more and better mitochondria and becoming more sensitive to insulin.
In older adults doing 12 weeks of aerobic training, the group taking metformin lost much of the gain in mitochondrial respiration that exercise produced, and the improvement in whole-body insulin sensitivity was blunted compared with the group on placebo.
In another trial of older adults doing 14 weeks of resistance training, the people taking metformin gained less muscle mass and less thigh muscle than those taking placebo, and the usual training response in their muscle tissue was dampened.
I'm not overtly opposed to metformin, and for a sedentary person with high blood glucose it can have a place, but I see its usefulness dropping as exercise goes up, since the more someone trains, the more the drug works against the very adaptations they're working for.
I also find it bizarre that metformin gets promoted as a longevity hack, because it's hard for me to believe that a drug that compromises mitochondrial function, which is so central to healthy aging, would be the key to living longer.
For anyone using exercise to improve their metabolic health, and especially for older adults trying to hold on to muscle, it's worth knowing that metformin and exercise can pull in opposite directions inside the muscle cell.
gronk 🇪🇺 retweeted
Sumerian erotic poetry: what it lacks in subtlety it makes up for in calcium.
Replying to @BamaExpat
Adonis’ Mid-Eastern equivalent Dumuzid/Tammuz has story of an underworld descent tied to changing seasons that looks more like Persephone than Adonis.
But the Greeks had an annual festival of weeping for the god’s death…and so did the Israelites, among others.
gronk 🇪🇺 retweeted
Reality of our days:
1. I am sitting on a set of vulns (4+) that allows me to remotely take over a consumer home automation device I have (3+ diff ways).
2. The vendor doesn't only serve consumers, but also critical infra (announced partnerships include pharma manufacturing plants, airport logistics, hospitals, etc) across various locations in Europe and the US (not the same product line, ofc, so maybe moot).
3. Claude and Astra refuse to help me explore even my own device, so now I'm using Chinese hosted models for investigating.
4. The vendor (with a revenue of $0.75B) does not maintain any security disclosure practice whatsoever, nor does respond to emails.
So, to recap, I need to outsource vulnerability hunting of (potentially, I don't know) critical western infrastructure to Chinese models, because the western models refuse; and the vendor doesn't give a shit. Fun timeline.
PS: The vendor is Swiss.
gronk 🇪🇺 retweeted
This may be one of the best explanations of male depression that I’ve ever heard. This is explains why guys are suicidal and their friends, parents, and/or wives don’t even know it. This is also why a man can be married with kids, great job, etc. & feel lonely & depressed.
Readers added context they thought people might want to know
This is an interview with @shaistaub who has not been credited for it. x.com/shaistaub/stat…
gronk 🇪🇺 retweeted
People who live past 110 carry a strange immune cell: a helper T cell that acts like a killer. It makes up about 18% of their T cells, over 4 times the share seen in people in their 70s to 90s. And its receptors match ones found in lung tumors. A new Cell Reports study asks whether these cells help explain how they got there.
Researchers from Keio University, RIKEN, and the University of Osaka profiled T cells from 28 Japanese donors: 8 in their 70s to 90s, 10 centenarians, and 10 supercentenarians (110 or older). They analyzed over 43,000 cells, tracking gene activity, surface proteins, and each cell's receptor.
First, a quick primer on the immune cells involved.
T cells are immune cells that patrol your body. CD4 T cells are usually helpers. They coordinate the immune response rather than kill anything directly. CD4 CTLs are an unusual subset that turn into killers, releasing toxic molecules like granzymes that destroy infected or abnormal cells. In most adults they're rare, under 5% of T cells.
Every T cell carries a receptor that recognizes one specific target. Cells that share the same receptor are a clone. When a clone gets repeatedly stimulated by the same target, it multiplies.
What they found:
• CD4 killer cells rose with age. Median share of T cells was 4.0% in the 70s to 90s group, 9.6% in centenarians, and 17.6% in supercentenarians.
• Helpers convert to killers in steps. Cells first lose a marker called CD27, then CD28. The in-between state (CD27 negative, CD28 positive) made up 10.5% of CD4 T cells, while the reverse order was under 0.3%.
• The cells weren't worn out. They stayed negative for PD-1, a marker of exhaustion, and had low levels of other exhaustion genes. They look highly specialized, not depleted.
• A few clones dominated. The single largest clone made up 33.3% of CD4 killer cells on average. In one centenarian, one clone was 126 of 234 cells. That pattern points to repeated stimulation by a persistent target.
• The dominant clones were private. No two people shared the same top receptor, which suggests each person's cells were responding to their own targets.
• One clone, many roles. After lab stimulation, cells from the same clone split into groups producing different signaling molecules, including types linked to antibody help, immune suppression, and allergic-type responses. All of them still made the inflammatory molecules IFN-gamma and TNF-alpha.
Then the cancer angle. The authors searched a database of over 690 million receptor sequences. Of the top clones that matched, 32 of the 36 expanded matches came from cancer patients, mostly lung cancer (17). But none of the supercentenarians or centenarians had a history of lung, breast, or liver cancer.
The team suggest these cells may be responding to something subclinical, such as early abnormal cells, senescent cells, or reactivated latent viruses. Earlier work showed CD4 killer cells can clear senescent cells in aged skin and suppress tumors in mouse models.
Some very important limits here:
Only 28 people were studied, and the cells came from blood, not tissues, so we don't know what they actually target. This study didn't test whether these cells kill cancer cells. The database matches rest on one receptor chain and show overlap, not proof of the same target. The cytokine results come from lab stimulation.
The authors also note these cells cut both ways. In other settings, they're linked to autoimmune disease, cardiovascular disease, and excess inflammation. Whether they help or harm may depend on how well inflammation is kept in check.
Still, it points to a possible feature of aging well: an immune system that shifts from broad coverage toward focusing on the threats it has already learned to recognize.
For longevity, the idea that immune aging is only decline may be too simple. These cells looked specialized, not exhausted. But these are people who already reached 110, so it's unclear whether the cells helped them get there or simply reflect an immune system that aged well. That second point is my inference, not something the study tested.
Immune function at 100 likely reflects decades of exposures, inflammation, and repair, and how well those were managed from midlife on.
gronk 🇪🇺 retweeted
>be comp biologist in 2005
>chimp genome just dropped
>everyone comparing genes, looking for protein differences
>nah, i have a better idea
>find DNA that mouse, rat, and chimp all agree on
>stuff that hasn't changed in 300 million years
>*ultraconserved*
>then check if it suddenly breaks in humans anyway
>run the scan
>49 regions light up
>only 2 of them code for proteins
>mfw the most human parts of the genome aren't even protein genes
>top hit: 118 bases
>chicken vs chimp: 2 changes in 310 million years
>chimp vs human: 18 changes in ~6 million
>name it HAR1
>human accelerated region 1
>cyberpunk name af
>"what does it do?"
>"we'll find out later"
>be wet lab biologist down the hall
>handed a 118-base mystery
>colleague in Brussels is staining embryonic brain tissue
>mail him a probe, "try this"
>months pass
>email arrives
>it lights up in Cajal-Retzius neurons
>7 to 9 weeks gestation, right when the cortex is building its layers
>it's an RNA gene, folds into a new shape in humans
>10 of the 18 changes keep the fold intact
>evolution was being careful
>still can't say what it actually does
>be enhancer lab in 2008
>run your own human accelerated scan
>top hit: HACNS1, another HAR
>stick human version into a mouse embryo with a blue dye reporter
>blue in the paw
>specifically where the thumb and wrist will form
>chimp version barely does it
>swap 13 human mutations into the chimp version
>now it does it
>we may have found part of the genetic story of the human hand
>still not sure of the general human accelerated function though
>be population geneticist
>see everyone celebrating "positive selection"
>notice the mutations are suspiciously biased toward G and C
>gc_biased_gene_conversion.jpg
>recombination can fake a selection signal for free
>build a model to separate them
>somehow the HARs survive the cut
>about a fifth of them are probably just gene conversion
>be the guy who brings this up at every conference
>nobody likes this guy
>but this guy is right
>be autism geneticist in 2016
>if these regions made the human brain
>breaking them should break something human
>check families where parents are related
>find HAR mutations disrupting both copies
>up to 5% of autism cases in those families
>some reviewers: "most of those CNVs hit genes too"
>shrug
>be machine learning guy, 2023
>train a model to read enhancers
>score every human-chimp difference in ~2,600 HARs
>expect all mutations to push the same direction
>43% of HARs have mutations fighting each other
>one turns the enhancer up, the next turns it down
>evolution overshot and corrected
>let's keep pushing
>huh, human DNA folds differently in the nucleus than chimp DNA
>structural rearrangements over a million years ago
>suddenly old enhancers are next to new genes
>"enhancer hijacking"
>post it, feels good, finally solved
>be yale lab, 2025
>map what 1,590 HARs actually touch, human and chimp side by side
>nearly 3,000 targets
>almost all shared between species
>no hijacking, just retuning old genes
>say the other guys were wrong
>post it, this is *not*solved
>be 2025
>we have CRISPR, organoids, and humanized mice
>put human HARs in chimp neurons
>they grow more branches
>put human HARE5 in mice
>more neurons, bigger cortex
>delete HAR123
>mice get worse at unlearning things
>finally testing them one at a time
>~3,000 to go
>be present day, present time
>telomere-to-telomere genomes for every great ape
>find whole new rapidly evolved regions that weren't even in the old reference
>call them HAQERs
>mfw more HARs
>the list keeps growing
>these things keep popping in every genome scan
>mfw the only thing accelerating here is our confusion
>mfw we still don't fully know what HAR1 does
>be cell biology postdoc at UCLA, 1986
>job: purify vesicles from rat liver
>ez
>purify the vesicles
>something else in the tube
>weird little barrels, way bigger than a ribosome
>fkisdis
>purify the barrels instead
>mostly protein, tiny bit of RNA
>not related to vesicles at all
>new organelle just dropped?
>lab holds a naming contest
>"vaults" "sure, "great name" "
>write it up, publish
>conclusion section: "function: unknown"
>surely we'll figure it out in a couple years
>reviewers: "are you sure this isn't just a virus"
>prove it isn't a virus
>apply for grants to study it
>first question on every call: "what does it do"
>"we're working on it"
>be early 90s
>vaults hang around the nucleus
>"they must plug the nuclear pores"
>they do not
>be different scientist, Netherlands, 1995
>chasing LRP, mystery protein that shows up in chemo-resistant tumors
>clone it
>~88% identical to rat vault protein
>it was vaults the whole time
>whole cancer field piles on
>a decade of papers on vaults and drug resistance
>nobody ever proves vaults cause it
>back at UCLA
>one of the vault proteins binds telomerase
>"guys what if this is the fountain of youth"
>knock it out in mice
>telomeres fine, aging fine, mice fine
>2002
>knock out the main vault protein
>mice: perfectly healthy
>knock out all three vault proteins
>completely vaultless mice
>mice: basically fine, some subtle immune stuff
>mfw something in every cell since the dawn of eukaryotes does nothing obvious
>"just test it in yeast"
>yeast don't have vaults
>"worms then"
>no vaults
>"fruit flies"
>no vaults
>mfw every easy model organism dropped them
>try making recombinant vaults in insect cells
>expecting to get some loose protein at best
>get fully assembled vaults
>can't find a single free copy of the protein anywhere
>turns out ribosomes 3D-print the barrel straight off the mRNA
>lolwut
>still no function
>be structural biologists
>spend years modeling it as 96 proteins with 8-fold symmetry
>2009 X-ray structure
>it's 78 proteins, 39-fold
>caps still blurry
>2025: caps are closed, 13-fold, tiny hole in the middle
>still no function
>NIH funds vault research for 15 years
>NIH: "so what does it do"
>"...we're working on it"
>NIH stops returning calls
>if we can't figure out what it does, we'll make it do something
>stuff a cancer-fighting chemokine inside the barrel
>inject into mouse lung tumors
>immune cells swarm in, tumors shrink
>2011: "human trials within three years"
>2024: "human trials next year"
>2025 database: preclinical
>start a YouTube channel explaining vaults to anyone who'll listen
>be different scientist, Broad Institute
>want to record what genes a cell turned on over time
>try tethering mRNA to a bacterial protein
>doesn't work
>stumble onto a YouTube channel about some obscure barrel organelle
>engineer vaults to grab mRNA
>RNA half-life goes from ~17 hours to ~132 hours
>publish in Science, 2026
>use it to catch drug-resistance genes in lung cancer cells before the drug even hits
>the vault guy is thrilled
>still nobody knows what vaults are for
>be epigeneticists
>screen the whole genome for methylation set at random in the early embryo
>separately study kids conceived in different seasons in rural Gambia
>both screens independently point at the same gene
>it's the goddam vault RNA gene
>its methylation depends on conditions around conception, locked in for life
>linked to cancer outcomes, blood sugar, cleft palate
>mfw the mystery barrel has a mystery RNA that remembers when you were conceived
>40 years later
>one of the most abundant structures in your cells
>still not in a single textbook cell diagram
>we use it anyway because yolo
>function: unknown
gronk 🇪🇺 retweeted
Declassified photos from Polish government archives show a team of safecrackers inside the vault of the US consulate in Poznan one night in the early 1980s.
The men were specialists from Section IX, Department II of the Secret Service of the Polish People’s Republic. Throughout the Cold War they regularly penetrated foreign embassies and consulates throughout Poland at night when they were unoccupied. The team didn’t just break in once but entered many of the supposedly secure facilities on a weekly basis, disabling security systems, picking locks, and photographing documents.
One of the keys to their success was also their greatest danger: radiation. Section IX built powerful, portable x-ray machines to bombard vault doors and reveal their inner workings. Team members were exposed to radiation 500x greater than was safe, and many succumbed to cancer later in life.
In the case of the consulate pictured here, a curved steel rod holding a radioactive isotope was inserted through a tiny hole drilled below the vault door. X-ray film was then placed in front of the Mosler combination lock, and the isotope would reveal the inner workings of the wheel mechanism within approximately 90 minutes, all while bombarding the Section IX men with radiation as well.
Once they were inside, the Mosler file cabinet was not bolted to the floor so it could be tilted forward and the isotope again placed behind it, and the x-ray film in front. They only needed to use these methods once until the combination was changed.
Section IX was so good at their craft that Polish records show they broke into the US consulates in Krakow and Poznan 44 times in 1982 alone.
For episode 122 of the Spycraft 101 podcast I spoke with Tomasz Awlasewicz, author of Invisible: The Greatest Secret of the Secret Services of the Polish People’s Republic. We discussed the methods used to access the most secure rooms of foreign embassies and consulates inside Poland, and the heavy price the technicians and operatives of Section IX paid for their groundbreaking work. The episode is available now.
#coldwar #consulate #safecracker #radiation #xray
youtube.com/watch?v=Xp0IbveH…
gronk 🇪🇺 retweeted
This is one confusing thing! In almost every way my dataset replicates obvious stats like this - women watch less porn, libs are more genderqueer, etc. - except for this. I find most gay men don't really have more sex partners, except for a small minority of them
gronk 🇪🇺 retweeted
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math.
rikiyat.github.io/blog/weight-sp…
github.com/RikiyaT/lm_und…
huggingface.co/Qwen/Qwen3.5-0…
huggingface.co/RikiyaT/Qwen3.…
gronk 🇪🇺 retweeted
Here are some random facts about your health that you might not be aware of:
-Kiwifruit supplementation improved vitality and mood within 4 days, peaking around 14-16 days
-Every 5-gram increase in daily fiber intake correlated with an approximate 5% drop in the risk of depression.
-Women who consumed less than the recommended amount of red meat were twice as likely to be diagnosed with a clinical depressive or anxiety disorder.
-Neuroscientists often can not distinguish between people with mild/moderate TBIs and people who have been exposed to mold
-Bibliotherapy (reading) is found to be highly effective for reducing long-term depressive symptoms in adults.
-Office job workers had higher levels of anxiety and depression than those working in manufacturing jobs
-People with major depressive disorder have serum zinc concentrations averaging roughly 1.85 µmol/L lower than healthy controls.
Why?
Well in a nutshell, zinc is crucial for maintaining homeostasis in BDNF, NMDA, GPR39, microglial cells, ACTH and more.
-Believe it or not, there's some research suggesting that botox helps with depression despite it being toxic.
The main theory is that the brain relies on a bidirectional feedback loop called facial feedback.
So when you feel sad, you frown for example.
However, if your forehead muscles physically cannot frown, the feedback loop to your amygdala is interrupted, telling your brain that the distress is less severe than it thinks.
So instead of botox, pay more attention to your facial expressions and artificially manufacture more positive ones :)
-When you are depressed, it's a good idea to pay attention to your sleep and circadian health.
But in the short term, staying awake for 36 hours straight has been documented since the 1970s as one of the fastest-acting antidepressant interventions in existence.
IF you have unipolar depression. If you have bipolar depression it needs way more caution (triggers full mania in 5% of people and hypomania in 6%) because well, this works mainly by rapidly increasing dopamine, norepinephrine and BDNF by downregulating NET for example (NET is responsible for sweeping norepinephrine out of your brain's synapses).
Note: It's not a long term strategy.
For roughly 80% of responders, the antidepressant effect disappears entirely after the very next period of sleep.
-Some of the benefits SSRIs might actually come from modulating ZnT-3, which is solely responsible for packing zinc into synaptic vesicles.
So they might force more zinc out of tight intracellular storage and into the active synaptic clefts of the prefrontal cortex.
-Slow COMT + mercury exposure = cooked brain (especially in children).
Mercury is well-documented to impair catecholamine degradation by interfering with S-adenosylmethionine (the methyl donor required by COMT).
This reduces COMT activity and thus leads to accumulation of norepinephrine, epinephrine, and dopamine.
The same SAM disruption can also impair phenylethanolamine N-methyltransferase (PNMT), shifting the adrenal output toward more noradrenaline and less adrenaline.
-NAC is very effective for compulsive disorders.
In a lot, if not most, compulsive disorders and BFRBs (hair pulling etc), there is dysregulation of glutamate homeostasis in the cortico-striatal circuits.
Mainly, the projection from the prefrontal cortex to the nucleus accumbens (NAc) fires excessively and this hyperactivity drives the "urgency" and loss of top-down behavioral control, making repetitive behaviors highly compulsive.
The extra cysteine from NAC stimulates the cystine-glutamate antiporter.
This increases extracellular glutamate levels outside the synapses/binds to metabotropic glutamate receptors(*), which shut down the hyperactive release of glutamate into the synapse via a negative feedback.
(*)primarily mGluR2/3)
BUT the same reason why NAC helps with compulsive disorders is why, if you use big doses for prolonged time periods, you will end up with anhedonia if you don't have one.
The same mechanism that dampens the "pathological noise" causing cravings or tics blunts the clean glutamatergic signals required for learning and reward processing for example. This was the short answer.
-Baicalin (a flavonoid found in Chinese skullcap) increased VMAT2 levels and enhanced dopamine balance in the brain.
-Plasma zinc and erythrocyte magnesium should be considered as biological state markers and used as such in both major depression and bipolar disorder.
-Ketosis might be a good additional tool to fight addiction more easily.
A ketogenic diet increases baseline levels of adenosine in the brain.
Adenosine receptors naturally pair with dopamine receptors to form functional complexes.
When adenosine levels rise, they directly repress hyperactive dopamine signaling.
Because cocaine's behavioral effects (like rearing and hyper-ambulation) rely on massive surges of dopamine, adenosine acts as a built-in brake.
On top of that:
-Ketosis increases the production of GABA
-β-OHB blocks the overactivation of an enzyme called CaMKII-α through a process known as β-hydroxybutyrylation.
By dampening this enzyme in brain areas like the hippocampus, the diet disrupts the structural "drug memories" and behavioral conditioning that drive cocaine addiction.
-A landmark 2019 follow-up study found that 44% of participating children no longer met the diagnostic threshold for autism spectrum disorder two years after completing a specialized form of fecal microbiota transplantation.
-Drinking at moderate levels (14–21 units per week (a pint a day)) tripled the odds of right-sided hippocampal shrinkage compared to abstaining.
Higher intake (over 30 units weekly) raised this risk even further.
The main mechanisms seem to include acetaldehyde and brain iron accumulation.
At the start of the study, 83% of the children were rated as having "severe" autism.
Two years after treatment, that number dropped to 17%.
-Participants who met the RDA for zinc, copper and selenium had significantly lower odds of depression.
-Aluminum inhibits protein phosphatase 2A, which helps keep tau and neurofilaments dephosphorylated.
It also downregulates TH-immunoreactive neurons inside the substantia nigra.
-Cracking up your window while you sleep might improve your brain health more than you think.
Cognitive scores drop roughly 50% at 1,400ppm compared to the 550 ppm baseline.
-Rates of depression correlate with the geographical distribution of deuterium in the natural water in the US population.
-Consuming a modified Mediterranean ketogenic diet reverses the peripheral lipid signature of Alzheimer’s disease in humans.
-Studies from the United Kingdom and Sweden indicate that the risk of depression is increased by at least 20%, with the former (over 1 million participants) reporting an increased risk with the number of courses and agents used, that persists with a slow decline over the ten years following exposure
-Inadequate dietary omega-3 intake after birth may similarly deplete maternal tissue reserves, potentially contributing to postpartum mood changes or cognitive fatigue
-In a study of 18 patients diagnosed with schizophrenia, 15 (83.3%) had positive serum FRα autoantibodies compared to only 1 in 30 controls (3.3%)
FRα autoantibodies can interfere with how B9 is transported across the blood-brain barrier into the central nervous system.
This blockage can cause a lack of folate in the brain even when blood levels of folate appear completely normal.
Some patients in related observations showed clinical improvement after receiving high-dose folinic acid (leucovorin), which can bypass the blocked receptors.
-Approximately 30% to 40% of patients with moderate/severe depression experience a meaningful improvement in symptoms with placebo and for individuals with mild-to-moderate depression, that response rate can climb to 70%
-The more vitamin D increases (up to a point), the less depression a person experiences.
-A high magnesium intake is associated with larger brain volumes.
-Glutathione levels are significantly reduced in the occipital cortex of patients with major depressive disorder (MDD).
-There's a higher abundances of certain bacteria like Firmicutes, Clostridia, Actinomyces, Bacteroides, Lachnospiraceae, Blautia, and Dorea are noticed in individuals with suicidal behaviors.
-In German rare-disease clinics, more than half of adults with unexplained symptoms had a current mental disorder and in almost 54% of those cases the mental illness alone explained the whole picture.
-Adolescents with social anxiety had a shifted microbiome (higher Prevotella, lower Parasutterella).
Transplanting that community into newborn rats produced anxiety-like behavior, social deficits, and metabolic changes in medial frontal cortex
-People with MDD often have 20–60% fewer butyrate producers.
-Stool from depressed humans, put into rodents, transfers anhedonia and anxiety-like behavior plus a twisted tryptophan metabolome
-Childhood unpredictability (chaotic routines, who shows up, what happens next) predicts depression even when ACE scores are zero.
-Ten years after clinical depression, about 10% of people are not just “not depressed”, they score above a quarter of never-depressed adults on well-being.
This means that recovery can overshoot baseline
-Suicide and depression rates climb with altitude in U.S.
-Youth cannabis use has been tied to higher later odds of schizophrenia (~+52%), recurrent MDD (~+30%), and anxiety (~+21%)
-Finnish men who sauna 4–7 times a week (vs once) had on the order of 65% lower dementia/Alzheimer risk and about 77% lower psychosis risk over decades
-One session of whole-body hyperthermia (core ~38.5 °C) produced antidepressant effects that were still visible six weeks later in a small RCT
-Antibiotics can be life-saving, but they are not Tic Tacs.
Do two or more rounds of them each year and your relative risk of depression increases by as much as 40%.
"I'll just take probiotics".
They can help but it's not how this works.
The effects of antibiotics extend far beyond simply lowering the count of good bacteria.
Classes such as fluoroquinolones and penicillins bind to and block GABA receptors in the brain for example.
-Impaired glucose metabolism is not only relevant for Alzheimer's when it comes to mental health/brain related issues but pretty much every single one because, well, the brain is a very glucose-sensitive tissue.
Areas like the hippocampus, amygdala, and prefrontal cortex rely heavily on GLUT4.
-Sleep irregularity is consistently tied to 26–53% higher dementia risk.
-Shift work is associated with 30–33% higher overall adverse mental health outcomes.
-Brisk walking for an extra 150 minutes per week is associated with living ~4.5 years longer on average.
-Social/friendship support interventions can lower depression susceptibility by around 30% in relevant reviews.
-Being insulin resistant increases your risk for depression by 2X.
-A single bout of RT or HIIT can increase levels of anti-cancer myokines and reduce the growth of MDA-MB-231 cells in vitro in survivors of breast cancer, potentially contributing to a lower risk of recurrence.
-Vitamin E therapy prevents hyperoxaluria-induced calcium oxalate crystal deposition in the kidney by improving renal tissue antioxidant status.
-CoQ10 levels in hyperthyroid patients were found among the lowest detected in human diseases.
-Lithium appears to be protective against aluminum toxicity in neurons.
-Creatine loading can increase DHT by 56% after 7 day and 40% if 5-10 grams are used after.
-You have a higher chance of being gay if your mom was hypothyroid during pregnancy.
-40g of dark chocolate can reduce menstrual pain as effectively as 400 mg ibuprofen.
-Colostrum is 3 times more effective than the flu vaccine.
-Night-shift workers show substantially higher rates of irritable bowel syndrome.
-When it comes to exercise a 2022 study saw the lowest risk of mortality at 150-300 min/week of leisurely vigorous physical activity, 300-600 min/week of leisurely moderate physical activity or a combination of both.
-There's a chance that we could achieve a COVID mortality rate pretty close to zero if everyone had a vitamin D level >50 ng/mL. Yes, zero.
-Vitamin D, A, B9 and zinc intake were lower in regions with the highest COVID-19 incidence and mortality.
-Choline is crucial for proper brain development and gestational choline deficiency in animals leads to lifelong cognitive deficits.
-Zinc at just 25mg can reduce depression by half and lower cortisol by up to 70% at 50mg.
-In rats, taurine protects their thyroid from fluoride exposure.
-Dietary vitamin B6 and vitamin E are associated with decreased odds of periodontitis.
-1–2 mg/day of elemental lithium (as lithium orotate or from mineral water) cuts suicide rates by up to 80% in high-lithium regions and reduces dementia progression in RCTs at 5 mg/day.
-Children with autism and/or ADHD seem to have impaired B6 utilization.
-High methionine, low folate and low vitamin B6/B12 (HM-LF-LV) diet causes neurodegeneration and subsequent short-term memory loss.
-Allithiamine can not only alleviate hyperglycaemia-induced endothelial dysfunction, but also reduce the level of advanced glycation end-products (AGEs).
-Rosemary can also reduce the level of advanced glycation end-products (AGEs).
-Astragalus supports mitochondria by scavenging harmful free radicals, reducing lipid peroxidation, increasing glutathione levels and it boosts endurance by 127% in mice.
-A combination of multiple lifestyle factors (sleep, activity, diet, social connection, etc.) can be linked to 50%+ lower depression risk compared with an unfavorable profile.
-Vitamin D dose-dependently reduces symptoms of depression.
-Cannabis is cooking your brain and testicles (CBD can be helpful though).
-Beta-caryophyllene (found in cloves and black pepper) increases salivary testosterone levels by 16%.
-Taurine is an essential neuromodulator during cortical development.
-Niacinamide can reduce plaque buildup in the brain and NMN improves insulin resistance.
-Children with autism often can not utilize folate properly due to genetic impairments.
-Not consuming enough B vitamins literally shrinks parts of the brain (especially B12).
-B1 is crucial for gut motility and stimulating digestive enzymes.
-High dose inositol (12–18 g/day) has antidepressant efficacy equal to SSRIs in multiple RCTs, with almost zero side effects.
-A copper:zinc imbalance (high Cu/low Zn) is found in A LOT of cases of postpartum depression and violent behavior in teens.
-10,800 FU of nattokinase per day showed significant reductions in atherosclerosis progression.
-Patients who passed away from COVID had low iron, vitamin B12 and vitamin D while also having high HbA1c.
-200 mg of vitamin C and 500 µg of B12 per day improve blood pressure, blood flow and reduce arterial thickness.
-Vitamin C lowers cortisol by up to 40% in some cases, increases oxytocin, supports cardiovascular health, increases dopamine and lowers prolactin.
-300 mg/day of ubiquinol (active CoQ10) reduces migraine frequency by up to 50–60%.
-Thiamine (B1) at high doses (up to 1.5 grams) cuts fatigue nearly in half in patients with MS.
-The first 27 longest living humans in history are women, and 80% of people over the age of 100 are women
-Plasma donation and blood donation have been seen to decrease the amount of forever chemicals in firefighters.
-Muscle androgen receptor density predicts muscle growth from resistance training far better than circulating levels of hormones
-PQQ seems good for elderly individuals with mild cognitive impairment.
-20 g/day of creatine loading + 5 g maintenance for 8 weeks cuts pain by 50% and fatigue by 60% in several small RCTs for fibromyalgia.
-High prolactin and low thyroid hormones are common in MDD.
-300 mg/day of ubiquinol cut cardiac death by 43% and total mortality by 42% over 2 years in patients with moderate to severe heart failure.
-High doses of biotin can lower blood sugar by half in diabetics and increase testosterone quite a lot (even by 3X).
-Vitamin C therapy might in fact increase the survival rate across all types of cancers by up to 4X.
-Vitamin A improves A LOT of symptoms of autism.
-Not getting enough selenium is so problematic that people with a baseline serum selenium concentration below approximately 106 ng/mL (within the low-normal range) had significant reductions in total cancer mortality and incidence of lung, colorectal and prostate cancers when supplemented with 200 µg/day of selenium.
-Separating fats and carbs in meals can minimize the simultaneous availability of both substrates, reducing competition and potentially improving glucose clearance.
-High dose riboflavin (400mg) works for up to 70% of migraineurs within 6 weeks.
-Vitamin A is superior to steroids for masculinization before 12y.o.
-10–15 g of glycine at night can improve insulin sensitivity the next day by even up to 50%.
-Men taking 600 mg/day of CoQ10 for 12 months reported a 113.7% increase in sperm concentration and a 104.8% increase in progressive motility.
-A single 20 g dose of creatine almost completely negates the cognitive decline from 24–36 hours of sleep loss.
-I can make you insane by changing your gut microbiome.
-Morning sunlight reduces the length of hospitalization in bipolar disorder.
-Religious people live longer.
-Higher vitamin B1, B6 intakes and plasma pyridoxal-phosphate are associated with lower risk of mortality up to 10 years.
-NAC + Glycine has been shown to expand lifespan in mice by more than 20%.
-Taurine has been shown to expand lifespan by 10% in male mice and 12% in female ones.
-At the end of a chromosome, there are these protective caps called telomeres and the consumption of bee products showed an increase of over 0.25kbp in just a year.
-Urolithin A is more effective than most of the supplements promoted for "anti-ageing" purposes.
-Group sports extend life expectancy and being a cardio freak (marathon maxing for no reason (run a marathon 1/year if it's for charity for example) shortens it).
-B12 supports ptp-3 and helps maintain good cognitive function in centenarians.
-Biomarker differences between centenarians and diseased populations include:
1. Metallothionein
2. Good 03:06 ratio
3. Glutathione
4. SOD
5. IL-6 (low)
6. TNF-a (low)
7. Uric acid (low)
-People who die by su*cide (but also have traits such as hostility(*)) tend to have modestly lower serum total/LDL cholesterol.
But lowering cholesterol with statins does not increase suicidal behavior.
So the proposed theory for this association remains that neuroinflammation changes lipid metabolism and results in these lower circulating cholesterol levels.
(*) patients with borderline personality disorder often have lower baseline fasting serum cholesterol levels compared to patients with other personality disorders.
-Eating breakfast is linked to lower suic*de rates
-A meatless diet increases the frequency of depressive episodes.
-When doctors say that depression is caused by low serotonin, they probably mean that it's caused by high quinolinic acid (its production outpaces its clearance, to be exact).
-Autism spectrum disorder seems to be the outcome of mitochondrial dysfunction during neurodevelopment.
-Magnesium intake is inversely associated with coronary artery calcification, with the risk being lowest at 450 mg/d of magnesium.
-Green tea enhances the skin anti-aging effects of red light.
-Giving hospitalized COVID-19 patients 30,000 IU of vitamin D daily for 3 days reduced their risk of death by 67%.
-Taking ibuprofen for more than 60 days in a year makes men twice as likely to experience infertility.
-66.5% of bankruptcies in the US are due to medical bills.
-Mold exposure can lower the IQ of children by 10 points.
-Vitamin D deficiency is linked to 60% higher odds of depression.
-Nattokinase can reverse carotid artery plaque by 36% at 10,800FU (taken over a year).
-Bright night light exposure is independently associated with type 2 diabetes.
-SSRIs can suppress melatonin by almost 50% under normal light conditions.
-Extending sleep by just 1.2 hours "offsets" 300 extra calories.
-The more you increase sodium fluoride, the dumber kids become.
-Exercise (aerobic) is more effective than SSRIS.
-It might sound weird, but onion juice is actually effective for alopecia aerata.
-Exercising 4 hours after reading something improves retention.
-Creatine is an underrated supplement for lowering triglycerides.
-Negative air ions (think being outside in nature) reduce depression by 50%.
-Thinking that depression is a response to inflammation might be correct in a lot of cases.
-By taking 600mg of NAC twice a day during flu season you might drop your chances of getting sick by up to 50% (note, NAC taken for weeks on end will deplete copper and harm some of the biofilms of beneficial gut bacteria)
-240mg twice daily of Fernblock may reduce the risk of sunburn by 400% and UV damage by over 300%.
-Bile treatment improves psoriasis.
-Using a smartphone for 60 minutes a day lowers fertility and testosterone quite a lot.
-Vapes can contain up to 20 times more lead than cigarettes.
-Laughter reduces cortisol levels by up to 37%.
-Vitamin K is associated with a reduced risk of cardiovascular, cancer, or all-cause mortality.
-Gingerol and shogaol in ginger enhance insulin sensitivity by increasing GLUT4 expression and inhibiting intestinal glucose absorption.
-Chromium enhances insulin receptor activity by improving the binding of insulin to its receptors and supporting glucose metabolism.
-Myo-inositol acts as a second messenger in insulin signaling pathways, improving insulin receptor function and glucose uptake.
-Fiber reduces postprandial glucose spikes.
-Fasting induces FGF21.
-93% of chronic fatigue syndrome patients had mycotoxins in their urine.
-Glyphosate was found in the sperm of 60% of infertile men in France (in the US a bit over 80% of the population have traces in their bodies).
-About 90% of people with fibromyalgia show elevated levels of mold-related toxins in their blood or urine.
-A higher 4-PA/PLP ratio (the marker of vitamin B6 turnover/catabolism) might be one of the strongest independent predictors of all-cause mortality in type 2 diabetics.
-One hour of chopping wood (with an axe) raises salivary testosterone by 48.6%.
-Two quite underrated yet effective tools when it comes to dealing with a candida overgrowth are undecylenic acid and baicalein.
-Tyrosine hydroxylase (the enzyme that converts tyrosine) can't function properly without thiamine pyrophosphate (TPP) as a cofactor.
-Lactoferrin (apolactoferrin might be even more powerful) helps against Covid quite a lot and is an underatted tool for lowering VC.
-630+808nm of RLT 2x/week for 6 weeks reduced abdominal fat by 4-5 cm in all participants without diet or exercise in a study.
-Quercetin and EGCG are zinc ionophores (they facilitate the transport of zinc into cells).
Hydroxychloroquine and ivermectin are also promoted that are zinc inophones in the alternative health spaces but that's not true.
-Giving 300mg of CoQ10 in fibromyalgia patients for 40 days led to 53% decrease in total symptom score.
-Taurine enhances IRS-1 tyrosine phosphorylation and reduces serine phosphorylation.
-Cold exposure activates BAT, which increases glucose uptake and fatty acid oxidation.
-A 25-year study of nearly 30,000 Swedish women found that avoiding sun exposure was a major risk factor for all-cause mortality (similar in magnitude to smoking). -The first time skin cancer was induced in mice was with a lamp that emitted UVC.
-Testosterone increases punishment of unfair behavior.
-Low creatine in the prefrontal cortex is associated with depression.
-A zinc deficiency does in fact harm androgens quite a lot.
-EGCG before doing cardio improves IR sensitivity by activating AMPK and reducing oxidative stress, promoting GLUT4 translocation.
-HIIT increases GLUT4 translocation.
-Light walking (15 minutes) after meals enhances GLUT4 translocation and glucose uptake in skeletal muscle, independent of insulin.
-Hypothyroidism impairs the activation and retention of riboflavin (and maybe iron, vitamin B12, folate, vitamin D, magnesium and copper as well).
*Thyroid hormones are necessary to regulate the enzymatic conversion of riboflavin (vitamin B2) into its active coenzyme forms: flavin mononucleotide (FMN) and flavin adenine dinucleotide (FAD).
That's it.
I ran out of space.
For more: fitandball.gumroad.com/l/bra…
gronk 🇪🇺 retweeted
On many days of the year, the River Danube simply vanishes.
Near Immendingen in southern Germany, the young river drains away through cracks in its own bed into the limestone below, leaving a dry, stony channel.
Where the water goes is fascinating.
1/3
gronk 🇪🇺 retweeted
There is a category of text message designed to force a response from your phone without ever alerting you.
In the cellular specifications (3GPP TS 23.040), it is formally defined as Short Message Type 0, commonly known as a Silent SMS or stealth ping. Inside the message header, the TP-Protocol-Identifier field is set to the value that indicates Type 0 (hex 0x40). The specification requires the mobile equipment to acknowledge receipt back to the network, but it must not indicate the message to the user, must not store it, and must not generate any alert (sound, vibration, or screen wake).
This happens because of the architectural separation in modern smartphones. The Application Processor runs iOS or Android, while a separate Baseband Processor runs closed-source firmware that manages the cellular modem. When a Type 0 message arrives, the baseband processes and acknowledges it at the firmware level. In normal operation it does not surface an event to the operating system’s telephony stack, so ordinary apps, notification settings, and location permissions never see it.
The silent acknowledgment is useful for location purposes. While a phone is idle, the network only knows which broad Tracking Area or Location Area it is in. Sending a Silent SMS forces the network to page the device. The phone responds with an uplink transmission; during that exchange the serving cell logs the Cell ID and measures Timing Advance.
Timing Advance is a distance estimate based on radio propagation delay. In classic GSM (2G) each step corresponds to roughly 550 meters. In LTE the granularity is finer (approximately 78 meters per step), and in 5G it can be finer still depending on the subcarrier spacing. A single measurement therefore places the phone in a ring around the tower rather than at an exact point. Investigators who can query multiple neighboring cells or send a sequence of pings can refine the estimate further and build a movement trail, all without any GPS data from the phone.
European authorities have publicly confirmed operational use. German parliamentary and government records have shown federal and state police, criminal police, and intelligence services sending large volumes of Silent SMS for location purposes, historically hundreds of thousands per year across agencies in peak periods, with individual state forces sometimes exceeding 300,000–400,000. Usage has varied by year and some statistics are now classified. The technique is also used alongside IMSI-catchers: the forced uplink burst gives direction-finding equipment a signal to home in on.
A VPN does nothing against this because SMS travels on the cellular signaling plane, not over the internet data path. Disabling location services only blocks apps from reading the GPS chip. As long as the cellular radio is active, the modem is required by the standards to answer network pages. The practical way to stop Silent SMS responses is airplane mode (or otherwise powering down the radio transceiver).
PS: Sorry, no simulator today :-|
gronk 🇪🇺 retweeted
We have discovered a substantial number of vulnerabilities in the German legal system using LLMs.
These may enable access to confidential documents, travel without an otherwise required visa, enforceable leave for soldiers at NATO’s JSEC, and more.
Similar to computer systems, the written law can contain vulnerabilities that invite exploitation. In our blog post, we share a selection of vulnerabilities, and present potential exploits for each. To our knowledge, none of these have been previously recorded. We share only those vulnerabilities whose exploitation we believe would not cause serious harm.
blog post: felixringe.com/legal-exploit…




































