Dan Patrick

Andrew Huberman just blew my mind on Modern Wisdom: You’re not a perfect 50/50 mix of your parents’ genes. Some entire brain areas can be 100% from mom… others 100% from dad. It’s not just mitochondrial DNA - certain structures that control hunger, behavior, and more can be almost entirely inherited from one parent. Human genetics is far more complex than the simple “half from each” story we were taught. It helps explain why some traits or tendencies feel so strongly linked to one side of the family. I’ve always found this kind of thing endlessly fascinating, it shows how unique each of us really is at a biological level. What do you think — does knowing some brain areas can be almost entirely from one parent change how you see family resemblance, or is this just another layer of the mystery?

People with low vitamin C levels burned 25% less fat during exercise. After vitamin C supplementation, fat oxidation increased nearly 4-fold.

Higher vitamin D levels in middle age are associated with less accumulation of tau, one of the pathological hallmarks of Alzheimer's disease. The strongest protection showed up among those with vitamin D levels at the higher end of the cohort (around 50 to 57 ng/mL), not just those avoiding deficiency. Interestingly, there was no association between vitamin D levels and amyloid burden. This doesn't mean vitamin D is a dementia-prevention drug, or that everyone should start chasing high levels. But it does reinforce that vitamin D status in midlife (and any age) is something worth paying attention to.

This works really well btw, at the end of your query ask your LLM to "structure your response as HTML", then view the generated file in your browser. I've also had some success asking the LLM to present its output as slideshows, etc. More generally, imo audio is the human-preferred input to AIs but vision (images/animations/video) is the preferred output from them. Around a ~third of our brains are a massively parallel processor dedicated to vision, it is the 10-lane superhighway of information into brain. As AI improves, I think we'll see a progression that takes advantage: 1) raw text (hard/effortful to read) 2) markdown (bold, italic, headings, tables, a bit easier on the eyes) <-- current default 3) HTML (still procedural with underlying code, but a lot more flexibility on the graphics, layout, even interactivity) <-- early but forming new good default ...4,5,6,... n) interactive neural videos/simulations Imo the extrapolation (though the technology doesn't exist just yet) ends in some kind of interactive videos generated directly by a diffusion neural net. Many open questions as to how exact/procedural "Software 1.0" artifacts (e.g. interactive simulations) may be woven together with neural artifacts (diffusion grids), but generally something in the direction of the recently viral x.com/zan2434/status… There are also improvements necessary and pending at the input. Audio nor text nor video alone are not enough, e.g. I feel a need to point/gesture to things on the screen, similar to all the things you would do with a person physically next to you and your computer screen. TLDR The input/output mind meld between humans and AIs is ongoing and there is a lot of work to do and significant progress to be made, way before jumping all the way into neuralink-esque BCIs and all that. For what's worth exploring at the current stage, hot tip try ask for HTML.

Brendan Fraser walked out of a Hollywood luncheon at the Beverly Hills Hotel in 2003, past a police officer in the lobby, and could not bring himself to say what had just happened to him. He went home and told his wife. He stayed quiet about it for fifteen years. The man who had groped him in that room was the president of the Hollywood Foreign Press Association, the group that ran the Golden Globes. In the years after, Fraser sank into depression, watched the work dry up, and wondered if the silence that followed was a coincidence. His body had its own bill coming. Fraser did most of his own stunts across three Mummy films. By the third one in 2008, he told GQ he was "put together with tape and ice." The surgeries that followed kept him in and out of hospitals for roughly seven years, including back operations, a partial knee replacement, and vocal cord repair. His divorce became final in 2009. A judge ordered him to pay 900,000 dollars a year in alimony plus another 300,000 in child support. By 2013 he was back in court asking to reduce the payments because he no longer earned enough. In November 2016, his mother died of cancer in Seattle after several years of treatment. His eldest son is autistic. Almost none of it made the news. In September 2022, his film The Whale premiered at the Venice Film Festival. The audience clapped for six minutes. Fraser sobbed on the balcony and tried to leave at one point, but the clapping wouldn't stop. In March 2023, at 54, he won the Academy Award for Best Actor. He thanked his three sons by name. The last years, he said, had felt like "a diving expedition on the bottom of the ocean," and his sons had been the ones holding the air line. Last week, he sat on Jimmy Fallon's couch and said he is training to play Rick O'Connell again. "I'm doing my best to get this 57-year-old gear in shape," he told Fallon. Universal has Mummy 4 penciled in for October 2027. He first played the character at 30, and he'll be 58 when the new one comes out.

Register here tinyurl.com/3x48dxv4

One of your most important parenting jobs is done by age 4.

Men, eat your carrots. Beta-carotene supplementation makes men’s faces look more attractive and healthy.

Looks insanely useful for the new voice api drop today

New Anthropic research: Natural Language Autoencoders. Models like Claude talk in words but think in numbers. The numbers—called activations—encode Claude’s thoughts, but not in a language we can read. Here, we train Claude to translate its activations into human-readable text.

Imagine taking your car to the shop and the mechanic says “you need your battery replaced” and you say, “how do you know that?” and he says “It doesn’t matter because I conducted a randomized controlled trial where replacing the batteries in half the cars on average made them perform better.” This is what “mechanisms don’t matter, outcomes do” should sound like to anyone with common sense. You really need to have common sense driven out of you relentlessly by education in order to arrive at the conclusion that you should entrust your health to someone who thinks how your body works simply doesn’t matter.

@yacineMTB @antirez @RepoPrompt

@antirez the problem is that they use a bag of things to measure impact, that bag of things is representative of the median user, of which you are not I think a neovim - tier tool for managing context is the end game where you explicitly add and remove files

This is the best way to impair your LLM agent, make it blind in order to save prefill GPU time and KV cache size (but actually not even the cache, the average KV cache of a session will always be 50% ctx_size on average). This sucks and is the path that turns great agents into bad ones.

What do you tell someone who's never going to run an ultramarathon but wants to make a change? @KenRideout_'s answer wasn't what I expected. He didn't talk about training plans. He reframed the whole thing around obligation — specifically, to whom you owe it. His argument is direct: your health isn't competing with your family responsibilities. It is your family responsibility. If you're making avoidable choices that degrade your health, and those choices eventually put a burden on the people around you — that's not a private matter. A parent is never going to be happier than their saddest child. And a child is rarely going to be at peace knowing their parent is struggling with something that didn't have to happen. That reframe carries more weight than the usual "you can't pour from an empty cup" language — which feels like permission. This feels like accountability. The actual starting point he recommends is humble: a one-mile walk. Then an hour walk. Maybe add a ruck. That's it. The goal isn't suffering for suffering's sake — it's the feeling on the other side of doing something hard. And that feeling, Ken argues, is available to anyone, every day. You just have to finish first.

GPT-5.5 + GPT-Image-2 is becoming one of the best combos for building apps! @dkundel breaks down why it works so well. We built those learnings into the Build Web Apps plugin, so Codex can handle the design-to-app loop for you. 👌

I was a little surprised by @sama's answer here. But if he’s right—and you can train a model completely on synthetic data—it would have all sorts of implications for how we develop and deploy this technology. Watch my full interview with Sam here: youtu.be/i9yXrdQ6noo Produced by @atlanticrethink, The Atlantic's creative marketing studio.

@jxnlco Enable gpt-image-2 to be able to produce authentic alpha layers in ChatGPT. You sound like the right person to advocate for it!

When I applied to OpenAI, I thought I would be working on evals. When I signed, I thought I would be working on agents. When I joined, I thought I would be working on Codex. After my first month, I thought I would be working on knowledge work, but here I am doing motion graphics.

Egg intake is associated with a lower risk of Alzheimer’s dementia.

Feels like everyone @OpenAI is using the Codex App & GPT-5.5 for everything these days, here's a few examples 1/ spreadsheets

Something Jocko said on a podcast I was listening to c. winter 2020-2021 changed my life—he was recounting how someone once asked him “what he says to himself” to get himself to do all the crazy disciplined stuff he does (up before 4am working out every morning, etc) and he was like that is the EXACT wrong question, you need to get out of the mind and into the body, you need to learn how to move the body by just going around the mind, let it scream and protest while you drag yourself out of bed, you cannot be held hostage by having to get the mind on board before you do anything