Nominee to lead FDA aims to speed up medical research, combat China’s rise statnews.com/2026/09/23/heid… via @statnews

My take on Galleri @WSJ @WSJopinion “People differ in how they value earlier detection, uncertainty, diagnostic risk and out-of-pocket cost. Once regulators have judged the evidence sufficient for access, informed patients and their physicians should be able to examine these results and decide together. Approval need not mean universal screening. It should mean that government doesn’t substitute one national judgment for millions of individual decisions.” @SylvesterCancer @asco @myESMO @SclcSMASHERS @FDA @LUNGevity @lungoncdoc @StephenVLiu @HHorinouchi @g_mountzios @sands_jacob @LauraAlderMD @oncodaily @OncoAlert @Latinamd @LuisPaz_Ares @RManochakian @LuisRaezMD @triparnasen @LaurenByersMD @LungCancerEu @alissajcooper @BZhangMD @NaglaAKarimMD @Annechiangmd @IASLC @NCCN @ASCO #WCLC26 @larvol @tribeMDUS @sylvestercancer @oncbrothers @chinmay @Jani_Chinmay @colazagasti @openmedicineHQ @myesmo @openmedkate @openmedben @chadinabhan @YoungLungCancer @OncodailyLung @ClinicalLung #LCSM @LungPolicy @dan_morgen wsj.com/opinion/the-galleri-…

This is insane. A young Iranian-American describes how his family has been hunted by the IRGC, cites tens of thousands of protesters reportedly killed by the regime, and asks Abdul one simple question: Will you condemn the IRGC? Abdul refuses.

If I were to be diagnosed, I’d *much* rather have it be at stage III than stage IV. The current difference in 5yr survival: 26% for stage IV vs 64% stage III (for all cancers), and gap widening with progress in therapeutics. The aggregate impact of 20% reduction on Stage IV in the US only is ~10.5K lives saved per year (with screening at scale in the eligible population). If you’d like to argue “it’s just lead time bias”. No: claude.ai/public/artifacts/4…

The NHS-Galleri trial results came out last Friday. 142,000 participants. 3 years. The most ambitious cancer screening trial ever run. The headlines: primary endpoint missed. That's a fact, and it matters. But I believe the headlines are missing the deeper story in these data – one that matters enormously to patients and to the future of cancer screening. Here's what the NHS-Galleri Test study showed: ✅ Stage IV cancer diagnoses – metastatic, typically incurable disease – declined substantially in the screened group, with >20% reductions across the 12 deadliest cancer types ✅ More cancers were found at Stage I & II, when treatment can be curative, in cancers that are typically diagnosed late (pancreatic, ovarian, liver, lung) ✅ 4x higher overall cancer detection rate when Galleri was added to standard screening (i.e., colonoscopy & mammography) ✅ Fewer cancers detected through emergency presentation – the worst possible way to learn you have cancer, and the most lethal and most costly ✅ No serious safety concerns. Test performance consistent with prior studies (with very high 99.5% specificity and very low false positive rate) So why did the 'primary endpoint' miss? See article for more detail & perspective on the study design, and the 'Stage Migration paradox' that masks enormous progress and clinical impact for patients.

Winchester 1873 for sure.

JUST IN: Idaho voters will choose the state’s official gun this November, with 6 firearms on the ballot. The Winchester 1873 is currently the clear favorite at 43%. poly.market/EC5b3h8

Today we announced the Claude-led discovery of a molecular machine that we suspect could represent a new gene editing mechanism. Its precise function, biotechnological utility (if any), or level of significance is not yet clear, but at minimum it is work I would have been proud to do as a PhD student. The work was done mostly, though not entirely, by Claude: our life sciences team suggested a broad area of research, Claude read through the literature and a bunch of genome data and discovered something interesting, then Claude proposed experiments to verify the discovery and our team carried them out. It’s easy to dismiss this as a one-off or curiosity, but we’ve repeatedly seen a pattern where AI performance in new intellectual domains goes from weak to superhuman in a matter of a few years. In 2023 models struggled to do math at the level of an average high-school student. In 2024 they started to do well on math competitions for the best high-schoolers in the country, in 2025 they started to solve minor open problems, in early 2026 more significant open problems, and in late 2026 they are beginning to solve the top few open problems in all of mathematics. We believe AI for biology is on a similar exponential trend. The main difference between biology and mathematics, of course, is that math can be done purely theoretically, while biology requires experimentation. Some have used this to draw the conclusion that AI’s utility in biology will be limited. We think this is wrong. As we’ve demonstrated today, humans can collaborate with AI to perform the experiments, validate key results in a few weeks and, if necessary, work with the AI to iterate on what they find. Eventually it may even be possible for Claude itself to safely perform the experiments by autonomously controlling lab equipment, with appropriate safeguards in place, but we aren’t doing that today (our lab is also a BSL1/BSL2 facility that doesn't handle materials dangerous to humans). More broadly, biomedical advancement has many stages — from fundamental biology discoveries, to translational research, to drug discovery, clinical trials, and finally the actual delivery of medicines and health care to patients. We are also interested in these later stages, but even simply accelerating the first stage of fundamental biological discoveries has the potential to speed up and broaden the entire pipeline. Improving our understanding of biology and sharpening biologists’ tools can drive forward all of the later stages, for example by identifying new drug targets, finding new therapeutic modalities, allowing for more precise measurement, and speeding up the experimental loop which itself further accelerates our understanding of biology. This will not in itself speed up clinical trial times, but if it succeeds it could greatly increase the number of promising candidates that go into the pipeline — an increase in throughput even though latency remains. In Machines of Loving Grace, I wrote about AI’s potential to “cure most diseases in 5-10 years” — a goal that sounds impossible, but one I believe is just barely possible if AI is applied to every stage of the pipeline. The first step is showing that AI can first help with, and then drive, biological discoveries. Claude’s discovery is the latest in a line of related prior work that goes back decades, beginning with systems like CRISPR, and continuing with discoveries like the bridge recombinase and VIPR in the past few years. Recently, there has been heightened interest in systems based on reverse transcriptase (RT) enzymes, the enzyme underlying the system Claude identified. And most recently, a Stanford team working independently described a novel RT system with an associated non-coding array that is in some ways similar to the one Claude found, though they are distinct systems that evolved independently from each other. I believe that we’re at the very beginning of finding such systems and developing them into powerful tools for biotechnology. I’m proud of the resources Anthropic has invested in accelerating the public benefits of AI through the life sciences, and we’re aiming both to grow our life sciences team and to work with other scientists to extend this approach to a broad range of problems. If you have a proposal for a research collaboration or are interested in joining our life sciences team, please reach out.

As we age, our blood quietly shifts toward making more inflammation-prone immune cells and fewer of the ones that fight infections and cancer well — real bad news in old age. This paper found one specific immune cell is secretly driving that shift, and blocking a single signal from it reversed the effect in old mice, restoring muscle strength too. nature.com/articles/s43587-0…

An FDA advisory committee reviews GRAIL's @GalleriTest multi-cancer early detection (MCED) test this Wednesday. The WSJ editorial board laid out the record this weekend. Lines worth reading: "Galleri caught four times as many cancers in total than the U.K.'s four recommended screenings for colon, breast, lung and cervical cancer." "More than half of cancers detected were at early stages, and 73% don't have recommended screenings." GRAIL's mission to 'detect cancer early, when it can be cured' approaches a big milestone that will drive down cost, make it accessible for many more, and save many lives. Let's hope the @FDA suports letting patients and doctors choose. wsj.com/opinion/a-grail-canc… @WSJEditorials @GRAILBio #DetectCancerEarly #MCED

Spoiler alert: you have two brains. The front of your brain and the back of your brain were never the same organ. They start as two different cells, in two different places, and they have been separate lineages for something like 550 million years. doi.org/10.1038/s41593-026-0…

Not only did more people die from overdose than Covid in San Francisco, the city spent $400 million dollars on shelter in place hotels for the homeless in which over 200 people died of overdose, plus hundreds of robberies and sexual assaults. Facts.

30 g of 85% dark chocolate daily for 3 weeks reduced negative mood and increased gut microbial diversity.

Ten women who had been breaking bones every year or two were given one infusion of their own marrow cells, and their fracture rate dropped by ~10X. A bananas result. All they had to do was give the cell with the right sugar coating. nature.com/articles/d41586-0…

In 1974, a performance artist in Naples, Italy stood passively for six hours while the ‘audience’ could use a range of 72 objects on her in whichever way they wished. She would not resist, for these six hours she’d surrender her own agency to the crowd and take responsibility for anything that happened. She’d remain completely passive. She left written instructions: “There are 72 objects on the table that one can use on me as desired.” And, “I am the object.” The artist was Marina Abramović and the piece was called Rhythm 0. But perhaps it is best viewed through the lens of a social experiment of sorts—decades before the MrBeast era. Of the objects provided there was quite a range. Some were benign: A rose. Honey. Bread. Apple. Comb. Lipstick. Cotton. Flowers. Mirror. Brush. Wine. Book. Perfume. Polaroid camera. Others, less so. Whip. Pocket knife. Chains. Nails. Kitchen knife. Saw. Hammer. Axe. Razor blades. And finally, a gun and bullet. At first, the audience was hesitant, gentle and playful. They offered her flowers, or wrote and drew on her body. They turned her position. But eventually things escalated. Once participants realized that there would truly be no consequences for what they’d do to her, they became more aggressive. They cut her clothes until she was half-naked, violated her bodily boundaries, sliced her skin until she bled, and pointed a loaded gun at her head. The person holding the gun tried to manipulate her fingers towards the trigger to see if she’d resist. It was aimed at her. It was at that point that some spectators intervened and took the gun away. Once the six hours passed, Abramović began to walk toward the crowd. Unable to face her, those who had been willing to touch, cut, strip, and injure her had literally fled the gallery. She was naked, bleeding, and crying. No longer just an object but a human, they couldn’t face her. It was a heartbreaking lesson about human nature. She said: “The experience I learned was that…if you leave decision to the public, you can be killed.” When you remove consequences, there will be people who will do the worst. “I really feel I was a mirror to the public, so however the public felt, it was projected on me.” But there’s another lesson here—perhaps a more significant one. It is about what happens when others remain silent. Most of the people who were there didn’t necessarily participate in the acts of cruelty. Most of them simply watched in silence. It was only when the situation escalated significantly that some people ended up intervening. But it was limited. After someone cut her neck, another spectator put on band-aid over the wound. When the gun was aimed at Abramović, several spectators physically confronted the person and took the gun away. That intervention possibly saved her life. But the silent played a role. As some became increasingly aggressive, their silence allowed them to continue. It was, in a way, perceived as permission. And a lack of social consequences. Once people began to act, they were able to prevent further harm to her. But the silence, one could argue, was part of the abuse. And why were they silent? Was it a desire to fit into the group? Conformity? Vicarious abuse and morbid curiosity? Fear of judgement? Not feeling responsible for the actions of others? A belief that someone else would intervene if things were really that bad? Support? We don’t know. But these are the same forces that can keep people silent when they witness all kinds of wrongdoing. We don't necessarily need to approve of something to allow it to continue as long as enough people stand there and do nothing. To me, it is a more interesting question than why people hurt her in the first place. Some people will always have it in them to hurt other human beings when presented with the opportunity. That’s the unsettling reality of human nature. But whether they are able to do so and how far they can go depends enormously on the environments around them. When people are silent. When they don’t intervene in the damaging actions of others. That’s when such wickedness is able to thrive. When everyone watches a disturbing thing happen, it gives it permission to continue. Silence isn’t neutral because of what it allows to happen. But when someone steps in, it often encourages others to do the same as well. They see that someone else sees that something is wrong and is willing to do something about it. Just as silence is contagious, so can courage be.

The definition of a broken regulatory system is when you can inject some random crap peptide cooked in someone’s basement tomorrow but if you have a deadly rare disease or a suicidal child, it takes years to get to you. We need to move faster. Faster.

We need smart/fast AI with no pause. Our adversaries will use a pause to win. Balancing the existential dangers is tricky but pause and losing to our adversaries presents dual risks. The “we lose” risk and “they still risk us all”risk. So there is one choice: we should go smartly forward. We do not avoid the existential AI risk by pausing. Unilateral Pause. Potential adversary win. Existential for us. Dual pause w adversary. Zero chance of compliance. They are losing and will do anything to win. No pause. Potential AI risk if improperly managed. This happens if either us or adversary wins and I trust us more to manage the risk (see the notes from the frontier labs and we do need some oversight). Potential losers will take greater not lesser risk. Adversary wins. Get both AI risk and world domination risk. Worst of all worlds. So smart forward only choice.

How to mitigate the Alzheimer's risk of APOε4? Block fibronectin accumulation. Fibronectin LOF gene variants protect the blood brain barrier and AD in APOE ε4 carriers. Work from the mouse model, postmortem human brains, clinical data, and induced pluripotent stem cells @NatureAging nature.com/articles/s43587-0…

BREAKING: Scholar Rock receives FDA approval for the first and only muscle-targeted treatment for all adults and children aged 2 and older living with spinal muscular atrophy (SMA) who are currently receiving a SMN2-targeted treatment. bit.ly/4gNc45r

The Fields Medallists are not ego-driven luddites, but they are human beings like everyone else, and they struggle to reason objectively when the social interest strongly conflicts with their status and sense of self-worth, and especially when this is jointly true of almost everyone in their social circle. It's the same reason to worry about Anthropic and OpenAI weighing private benefits against risks to everyone on earth. The actual argument in the Fields Medallists' statement is weak. It is undoubtedly true that famous problems have historically served important roles in research mathematics: both wrong turns and breakthroughs have led to the discovery of new ideas, and solutions have alerted other mathematicians, "This is something worth studying carefully." If mathematical training and institutions were forced to work the same way they worked historically, and if there were no countervailing benefits, then one can imagine a world where knowing the solution to famous problems without a clear explanation would be counterproductive. Except: mathematics is not frozen in calcite -- mathematical researchers can reallocate their time to different tasks -- and there are huge countervailing benefits. Students and mathematicians will need to allocate more time to explaining AI proofs, as many are already doing. And the role of human mathematicians will change further still as AI becomes good at explanation. We will need to find ways to insure students against the substantial risk they face. But most importantly, it actually matters to know the solution to problems, even if this solution is explained imperfectly. Knowing what's true is how you know what to build on, and how you build intuition to search the space of possible claims for further true claims. A physicist who was simply "told" correct equations for quantum gravity with no explanation might lose the joy of discovery, but this is indisputably progress in physics, even if other problems remain open, and even if the physicist must now question whether their efforts on those other problems will be necessary. Cancer cures, more efficient methods of chemical synthesis, or formulas that reliably predict earthquakes are undoubtedly progress, even if they put some people out of work and de-motivate others. Again, objectivity is extremely difficult when your livelihood and status are at stake. This is not an insult to anyone, it is a description of the human condition (or at least, the human condition so far...).