Sorry, friends, for digging in so hard on this topic, but it’s extremely important, IMO. I’m challenged on it constantly by academics who are bent on trying to get everyone else to believe the crap they had to memorize in order to gain their credentials. According to some of these people, I follow a “restrictive diet.” I’m still trying to determine precisely what I’m being restricted from. Oreos? Doritos? Added sugar? Foods I once had difficulty controlling—but no longer particularly want? If that qualifies as restriction, then I suppose I’m also on a cigarette-restriction diet, a heroin-restriction diet and a drinking-motor-oil-restriction diet. Academics can search PubMed and produce tens of thousands of papers containing the term restrictive diet. That establishes that the term exists. Congratulations. It does not establish that everyone who eliminates certain foods experiences restriction, deprivation or suffering. PubMed can identify what a diet excludes. It doesn’t typically tell you whether the person eating it feels restricted. That distinction matters a lot in the real world. If you sit down to meals you genuinely enjoy, eat until comfortably satisfied, experience no meaningful hunger or cravings, and feel healthier, stronger and more energetic than you did before, where exactly is the deprivation? Conversely, if you spend every day thinking about everything you “can’t” have, bargaining with yourself and waiting for your next opportunity to escape your diet, that sense of restriction is very real, even if your officially approved meal plan includes every food group. A dietary pattern can be technically narrow yet psychologically freeing. It can also be technically flexible while leaving someone hungry, preoccupied with food and miserable. That may be inconvenient for people who prefer labeling diets to listening to the human beings eating them, but inconvenience does not make it untrue. Years ago, avoiding added sugar would have felt restrictive to me because I still wanted it. Today, added sugar is so far outside my consideration that not eating it requires approximately the same willpower as not eating the cardboard box it came in. I don’t wander the supermarket mourning the Oreos I have been denied. I’m not staring wistfully at bags of Doritos, grieving the cruel limitations imposed upon me. I eat food I love. I eat plenty of it. Then I move on with my life. Call my diet restrictive if a journal’s terminology requires you to. But if I don’t feel restricted, deprived or hungry, your label tells me far less about my life than you seem to think it does.

🚨🚨🚨🚨🚨WHOA!!!🚨🚨🚨🚨🚨 Our film, tcc #TheCholesterolCode just got picked up by @PrimeVideo!!! It's now part of their catalog and you can watch it free -- right now! 👇👇👇👇👇👇👇 amazon.com/gp/video/detail/B…

Radiologists feel the need to remind me that dental imaging is different than other medical imaging, and that it somehow discounts the incoming wave of AI in all aspects of radiology. I don't really care what doctors think about AI's involvement in radiology. It's an objective assessment, and not even some weird prediction I'm making. Look around, ask questions instead of having (embarrassing) knee jerk reactions. But, I know you won't so I'll spell it out for you: AI is not something that comes knocking on your door and says "Hey Dr. Syed, you can go home now. I've got it from here." AI is something that spreads like cancer. First it shows up in your tools, then your EMR, and even on your radiographic machinery itself. Before you know it, AI is everywhere. Remember how pharmacists used to have agency in their profession? And then the PBM's/CVS turned them into glorified cashiers? That's what will happen. You will slowly lose agency at your job, and it will happen like boiling a frog (This is good and bad, and not exclusive to radiology!). Let's take a look at the numbers for those in denial: In 2020, ACR did an annual survey (sciencedirect.com/science/ar…) and found that 30% of radiologists were already using AI in their practice, and to no surprise, bigger practices were more likely to adopt it. That was 6 years ago. Of those 70% that didn't, 20% said they plan to in the next year. Ok, now let's get a little more current. The FDA *currently* lists 1,600 AI-enabled medical devices, and 75% of them are in radiology/imaging. Imagine how many are going to become listed in the next 5 years? GE Healthcare has: - spent over $700 million on R&D so far this year - 132 U.S. FDA-listed AI device authorizations - Completed $2.3 billion Intelerad imaging-software acquisition in March Siemen's: - Spent over 1.5B in R&D so far this year - In May, it announced FDA clearance for six interventional imaging systems incorporating AI image processing So yeah, Radiology and AI are about to become inseparable if they haven't already. And having a doctor need to fail at 10,000 chest x-rays in order to get good will be as primitive as the first EKG which required 5 people to operate and put the patient's limbs in buckets of saltwater. Even if you, the doctor doesn't use it, it's in your machines, it's on your computer, it's everywhere. It's insane that so many of the docs publicly admit that they cannot forecast what things will look like in their own profession in 5 years. Either that, or the reading comprehension is so low, that they can't properly evaluate AI without feeling insecure. That's problematic for an entirely different reason.

Riboflavin (vitamin B2) is a cheat code for migraine prevention. At 400 mg/day, randomized trials suggest it can meaningfully reduce migraine frequency. The mechanism is fascinating: riboflavin is a precursor to FAD and FMN, cofactors your mitochondria need to produce ATP and antioxidant defense. Migraine brains appear to have impaired mitochondrial energy metabolism, so B2 may essentially raise the brain’s energetic threshold for triggering a migraine and reduce neuroinflammation.

This is intended to increase the perceived value of doctors, especially radiologists. What it actually does, is quietly admit that the first 10,000 chest X-rays analyzed by a doctor shouldn’t be trusted due to the difficulty. Can’t think of a better AI endorsement.

at this point it’s unclear to me whether google can actually execute on the personal agent product it desperately needs as well as meta has or even openai. this category requires a very specific product dna which meta spent an enormous amount of money recruiting & acquiring. you need to collapse absurd technical complexity into something simple, friendly, opinionated, & intuitive then tell a compelling story. this is a non trivial task. also google’s other problem is that its product surface is so fragmented that i don’t even understand how the various orgs negotiate ownership, distribution, etc. i.e does thus live in gmail or gemini or both? ironically having just the google connector makes the problem much easier for everyone else. i genuinely can’t remember the last time google shipped a consumer product built from the ground up this complex that felt completely coherent end to end. anyway there should be a four alarm code red inside the company esp given how quickly meta can turn muse into a household name.

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.

Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages. Beside the enzyme’s gene sits a long array of repeating DNA—a structure that looks somewhat similar to CRISPR. We don’t yet understand what this system does, but only a handful of known systems share its features, and all of them are able to cut, copy, and paste DNA. Historically, the discovery of such programmable systems has helped revolutionize medicine. CRISPR, for instance, is now the foundation of genetic medicines. But it will take much more work to learn what this system does, and whether it can be put to similar use. Read more: anthropic.com/news/claude-di…