Men spend more per Botox visit than women. Had no idea. Also learned there's an actual term for it... Brotox. A few years ago I asked my wife if she thought it was important for men to 'look young' + get Botox like women do. She said absolutely not. She finds the plastic look severely unattractive + thinks the natural lines/wrinkles give men that rugged look as they age. Fair enough. Brotox avoided. Although if it ever takes 3 strokes off my golf game... we may need to revisit. The more interesting part to me is WHY ppl do this stuff. I've come to believe men/women aren't augmenting their appearance purely to attract the opposite sex. A lot of it is simply wanting to feel better when they look in the mirror. And if a little Botox genuinely improves someone's self-esteem, I'm all for it. It's your life. Who gives a hoot what anyone else thinks? wsj.com/health/wellness/why-…

Two ppl live from age 30 to 70. One biologically needs 6 hours of sleep. The other needs 8. Assuming both are equally healthy and fully rested, the difference is pretty insane. -- 14 additional waking hrs/week -- ~61 hrs/month -- 730 hrs/year -- 7,300 hrs/decade -- 29,200 hrs over 40 years That's 3.3 YEARS of additional conscious life. Was listening to Patrick/Huberman pod about energy + why certain ppl seem to have an entirely different gear. Huberman brought it back to something basic. Understanding how much sleep your body actually needs to operate at its best. Over the last ~5 years, I've found that my personal sweet spot seems to be around 6 hours. Getting 8 is impossible. But feeling rested obviously doesn't prove my body only needs 6. What's fascinating is that natural short sleepers actually exist. Rare genetic differences allow some ppl to function on significantly less sleep w/ out the typical consequences of sleep deprivation. For everyone else, sleeping less isn't some magical productivity hack. There's a real biological cost. But the thought experiment is wild. Two ppl with the exact same lifespan, yet one potentially experiences YEARS more waking life. Makes you wonder how much of what we attribute to someone's insane energy, drive, or ambition is simply biology giving them more usable hours.

The HOLY GRAIL of consumer FinTech will be a product you barely need to open. I've always been bullish on beautiful UX/UI. But the more I think about personal agents, the more I wonder how much of the traditional FinTech interface we'll actually need. Today we set up recurring investments/cash transfers/ automatic bill payments, etc. Pretty static/trivial stuff. Imagine an agent that knows your income, spending habits, NW, investments, goals, upcoming obligations, + how you behave with discretionary money. -- Every paycheck, it allocates your money across investments, savings, bills + discretionary spending. -- It anticipates large expenses + adjusts your cash position accordingly. -- You're having a kid, buying a house, or changing jobs? It adjusts alongside you. -- It notices shifts in your spending behavior + proactively helps you course correct. And here's the kicker. It doesn't need to be some superintelligent financial genius. It needs to be RELIABLE. We're talking about ppl's money. Trust is EVERYTHING. Over thousands of small, correct decisions, the agent earns your trust. Maybe eventually you don't think twice about letting it move $50K between accounts. We've spent decades building interfaces that help ppl manage their money. What happens when we build something that actually manages it for them? You still have control. You can see what it's doing, understand why, + intervene whenever you want. But you don't need to think about it every day. Go build your career. Raise your family. Take risks. Your personal FinTech agent GOT YOU.

Some of the best books I've ever read are about ppl who became obsessed w/ understanding something the rest of the world couldn't. 15 books I'd recommend to just about anyone: -- American Prometheus - Kai Bird + Martin J. Sherwin -- Genius - James Gleick -- Surely You're Joking, Mr. Feynman! - Richard Feynman -- Shadow Divers - Robert Kurson -- Rocket Men - Robert Kurson -- The MANIAC - B. Labatut -- When We Cease to Understand the World - B. Labatut --The Idea Factory - Jon Gertner -- A Mind at Play - Jimmy Soni + Rob Goodman -- Leonardo da Vinci - Walter Isaacson -- The Code Breaker - Walter Isaacson -- Genentech - Sally Smith Hughes -- Einstein: His Life and Universe - Walter Isaacson -- The Wright Brothers - David McCullough -- The Making of the Atomic Bomb - Richard Rhodes Would love to find 5 more books at this level. What the hoot am I missing?

The bigger story imo... or more of a detail for the nerds. Looks like Zeus (Zuck) named his AI agent Agrippa. Marcus Vipsanius Agrippa was Augustus's best friend + top general. Won Actium. Built the Pantheon. Basically a legend who ran the empire so Augustus could be emperor. That's the whole pitch for personal AI in one name. Not a chatbot... a right hand. I see you, Zuck.

When I was a v green analyst, one trait always stood out in the best operators around me. Title didn’t matter. They were religious about the follow-up: -- Best notes -- Key decisions -- Who owns what next -- Sent immediately It sounds stupidly simple. V few ppl do it consistently. Now I use @Granola for basically every meeting. AI has largely solved the note-taking part. But there’s still tremendous alpha in being the person who takes the AI notes, cuts the 90% bs noise, + sends: here’s what we decided + here’s what happens next. Follow through + closing the loop >>>>

X doesn't have a Wrapped. So I pulled my archive + built one. My account is 6 years old... I've posted on 297 of those days. So really... less than a year of tweeting. 741 posts. Posts per year: 2020 ▇▇▇▇▇▇▇▇▇▇ 209 2021 ▇▇▇▇▇ 98 2022 ▏ 8 2023 0 (thriving) 2024 0 (also thriving) 2025 ▏ 3 2026 ▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 423 I went 1,033 days w/out posting a single thing. Then 2026 hit + apparently I had a lot to say. 57% of everything I've ever posted happened this year. What changed (share of posts touching each theme, 2020–22 → 2026): -- AI + agents: 1% → 36% -- Product + building: 7% → 41% -- Money + markets: 9% → 23% What stayed the same: -- Mindset + quotes: 28% → 29% -- Film, TV + books: 16% → 15% -- Fitness + sports: 10% → 10% My voice changed too. Average post went from 21 words to 78. -- 2020 me collected other ppl's wisdom. -- 2026 me argues w/ strangers about agents. Also, 'fkin' has been my favorite word since I was a teenager. Used it 0 times from 2020–22. 36 times in 2026. More than I said 'Claude.' Can't stop now just bc I'm in my 30s. Same brain. New obsession. X should ship this natively. Until then, download your archive + go look. I plan to be MUCH more consistent over the next decade+. Bold claim from a guy w/ a 1,033-day gap. What's the biggest shift in your posting over the years?

Random thought this am: The fallacy of fitness... the ppl who look the most disciplined often need the least discipline to work out. Once you’ve trained long enough, exercise becomes a need. You feel like shit when you don’t do it. Going every day isn’t some heroic act of willpower anymore. The harder problem is the person w/ a brutal job, kids, commute, whatever, who has mentally priced “working out” at 60-90 mins. After a 20 min lift this morning... I think that price is wildly wrong. Most of us aren’t trying to be bodybuilders. We want to feel mentally sharp, stay kinda strong + maybe build a pretty solid physique. The math: -- 6 exercises -- 3 sets each = 18 sets -- ~30 sec/set = 9 mins lifting -- ~20 sec between sets = ~6 mins -- ~3-5 mins changing weights / moving around Call it ~20 mins. Add a 5 min warmup + 5 min stretch and you’re fkin out in 30ish mins. Get in. Work hard. Sweat. Gtfo. Go live your life. Makes me wonder how many other things I’ve turned into 90-minute problems before ever testing whether they were really 20-minute ones.

Pretty clear that AI is making it possible to do 10x more things at once + right now, breadth is getting rewarded far more than depth. You can juggle 10 projects, ship 10x more code, launch more products, + constantly move on to the next thing. It feels like having superpowers. But when are ppl actually going deep? Spending hours understanding a problem, connecting dots across domains, developing real expertise, or just sitting + thinking? I have strong conviction that long term, the ppl who can focus + go 10x deeper into specific domains will have an enormous advantage over those spreading themselves across 100 different things. That might not be what's rewarded today. Speed + sheer output get most of the attention. And that's the juggling act. You'd be crazy not to take advantage of what AI makes possible. But equally crazy to sacrifice depth in the process. Just bc you can do 100 things doesn't mean you should.

You can copy every single feature of a great product + still build something nobody gives a shit about. Think about Equinox v the YMCA. Both have weights, treadmills, locker rooms, showers. More or less the same feature set. But anyone who's spent time at Equinox knows the difference. The cleanliness, the mirrors, the lighting, the staff, the steam room, the shower after your lift. All these tiny details compound into a feeling. I've happily paid $200+/mo for Equinox. + only part of that value comes from the actual workout. This is how I think about building consumer products. Anyone can remove steps from an onboarding flow. Anyone can add beautiful animations or haptics. That doesn't mean you've created a great experience. The best PMs + designers go much deeper. They obsess over every single moment a customer interacts w/ the product. Not just whether the flow works, but how it makes someone feel. What emotion are you creating? How do you want someone to feel when they put your app down? As AI makes building software increasingly trivial, I think this becomes an even bigger differentiator. Features have ALWAYS been copied. That feeling though... is so much harder to replicate. PS... I tweeting this on the treadmill at the YMCA rn. Only reason I ever joined EQ was bc I lived next door to one in NYC. Accidentally developed expensive taste in gyms lol.

Love how easy these articles are to digest. Can't say my writing is there yet... but a man can dream. I digress... Thinking on "Have Complete Thoughts": The danger w/ complete thoughts is turning a great idea into an excuse to overthink. If you’re Apple making a massive hardware/software bet, yeah, you better think deeply about product, distribution, business model, ecosystem, GTM, + what winning actually gets you. The cost of being wrong is enormous. But a startup searching for PMF is fundamentally different. You cannot sit in a room + think your way into PMF. The market is part of the thinking process. Ship. Learn. Iterate. Let reality complete the thought. Maybe the rule is... the more expensive, irreversible, + ecosystem-dependent the bet, the more complete the thought needs to be before you leap. But when the cost of being wrong is low, sometimes the most strategic thing you can do is just ship the damn thing.

The moment there is a leaderboard… is the moment we’ll start seeing legends go pint for pint. Cheeky Pint podcast soon to have the best content on the internet come EOY 2027. The next chart will be the intern checking… at what pint # did things get really interesting on the pod.

Truly wild. Excited to see the incoming NFLX doc.

Best tweet I’ve seen all week. Directionally, I think he’s right. Timeline? No fkin chance. Few one-liners I'm thinking about: - Thinking + acting like an owner makes your work better - Building someone else’s dream gets exhausting - Building your own thing can be far more purposeful - Ppl want agency, freedom + real incentives But I think all of that is disproportionately true for high-agency ppl. There are millions of ppl who really just want to be told what to do. And thats OKAY (I think). I have friends in extremely high-caliber jobs who have told me, “I can’t work from home. I need my boss there to motivate me.” That sounds insane to me. But it’s v real. We condition ppl this way from grade school onward. Show up here. Do this assignment. Meet this rubric. Wait for the next instruction. AI can make the one-person company economically viable in 5 years. That doesn’t mean human behavior + work culture change in 5 years. 10-20 years feels much more believable to me. Hope I’m wrong.

Watched The Secret in Their Eyes (2009) this past wknd. What a beautiful film. It weaves past + present around a murder investigation, but the real brilliance is the deeply intimate relationship between two ppl who can't bring themselves to say how they feel. Emotionally agonizing to watch. As the past slowly explains the present, the film delivers multiple holy shit moments. Just masterful storytelling. 99% of ppl sleep on international cinema. There is SO much incredible filmmaking out there if you bother to look. imdb.com/title/tt1305806/

Some of these AI retention takes are getting so click-baity. Especially from ppl who should know better. Muse has incredible D1 retention. But claiming it's "almost as good as ChatGPT" after 14 days?! Come on. Let's look at the numbers pls. Assume 1,000 new installs for each app: Day 1 retention: - Muse: 73.3% (733 users) - ChatGPT: 41.3% (413 users) Day 14 retention: - Muse: 22.5% (225 users) - ChatGPT: 26.8% (268 users) Muse starts with 320 MORE users returning on D1, yet ChatGPT overtakes it around D9 and retains ~19% more users by D14. Anyone who's spent time building consumer products knows D1 matters. It's an early signal of activation, onboarding, curiosity, etc. But it tells you v little about whether you've built something ppl need over time. And that's the part we're missing! - What's retention at D30? - What about D60? D90? - Does Muse's curve flatten, or does it keep bleeding users? - How much of that initial spike is real utility vs. launch hype/novelty? rn, I'd assume it tilts heavily on the latter Also, why tf is Google Maps on this chart haha. Retention means different things depending on how frequently someone naturally needs the product. I'm bullish on personal AI. But 14 days of retention data is not enough to declare a durable consumer winner. Looking forward to combing through these numbers at the end of Q4.

I think a lot of very intelligent ppl are about to discover that they've spent their entire lives getting good at producing answers, w/ out ever developing a deep understanding of the things they're working on. Loved this essay on the technical debt of AI-generated mathematics bc the argument extends to practically every domain. We're entering a world where anyone can produce sophisticated-looking code, research, mathematical proofs, or product strategies in seconds. But producing something + deeply understanding it are two entirely different intellectual exercises. And when everyone can push the same button, producing the output itself is no longer alpha. I remember this vividly from engineering undergrad. Plenty of ppl figured out how to get straight A's. They knew exactly what to study and how to consistently produce the correct answers. But I also knew ppl who were mediocre test takers + brilliant engineers. Someone might screw up Ohm's Law on a circuit analysis exam, then spend an entire wknd designing a PCB bc they were fascinated by how the damn thing worked. I always admired the latter group. They weren't optimizing for the answer. They were trying to understand the underlying system. Claude Shannon is perhaps one of the greatest examples of where this type of thinking can lead. His deep understanding of information helped lay the foundations for modern computing + digital communication. We've always rewarded output bc it's easier to measure than understanding. Grades, publications, lines of code, features shipped. LLMs are exposing this problem at an extraordinary scale. If we can generate hundreds of mathematical proofs but nobody has the time to verify, explain, or connect them to broader human knowledge, what exactly have we accomplished? I believe strongly the real alpha will belong to ppl willing to do the mentally exhausting work of developing understanding. The ones who can explain something to a third grader, question an assumption everyone else accepts, or connect ideas across seemingly unrelated domains. Generating an answer is fkin trivial. Understanding why it matters has always been where the gains are at. terrytao.wordpress.com/2026/…

In 2015, @jockowillink went on the Tim Ferriss podcast + told a story that I think about constantly. He'd sit in meetings where his Navy SEAL Commander would go around the room asking everyone what they needed. Everyone would chirp something like... more equipment. more resources. more support. When he'd get to Jocko? "We're good, sir." Fkin love that line. For months my CEO has asked me some variation of "What do you need? Another eng? Another FTE?" And my answer is almost always the same. "Nah, we good." I want to stretch myself. I want to stretch the ppl I work w/. I want us juggling so many balls, operating right at the precipice of what we're capable of. Working so fkin hard that we're forced to find another gear. This isn't for everyone. I get that. But I believe ppl who can sustain that level of intensity, focus, + output can compress YEARS of learning into months. Rn we're a tiny product team. PM, designer, eng. We're all designing, building, testing, shipping production code. Every single day. It's remarkable what three ppl can accomplish now. For decades, the game in tech was to build bigger teams, get more headcount, manage bigger budgets. I'm increasingly obsessed w/ the opposite. How much can a tiny team of absolute killers accomplish w/ a fraction of the resources? And btw, I'm not comparing the team sport of Product Management to Navy SEAL operations lol. I'm fully aware we're moving pixels around a screen, not clearing buildings in Fallujah. But that mentality? I think it applies to just about any domain.