Matt Hebard

This is remarkable. Watch every minute. Movement truly is medicine. @tferriss m.youtube.com/watch?v=FcQ3EL…

When in doubt... let multiple agents run in parallel :)

So aligned here. Been on team HTML for months. So much easier to digest/present across a team.

Indiana high school athletes can now be paid NIL compensation. Three states still don’t allow high school athletes to receive NIL compensation, and Texas has a very limited exception.

I'm not a father yet. But I know how I'll feel when I am. Nick Maggiulli just wrote a great piece. His wife goes into labor at 1:30am. Heart rate drops. Half the hospital floods the room. Everything teeters on the edge. Then it stabilizes. Their daughter is born three hours later. He writes that becoming a parent gave him a deeper understanding of what investing actually means. Investing as in: every sleepless night, every sacrifice, every dollar worked for... placed in service of someone else's future. I've heard this from multiple fathers. That something sharpens when you're no longer building for yourself. The noise quiets. The priorities crystallize. The small stuff stops feeling like the main thing. I'm not there yet. But I know it's coming (..knock on wood)! And pieces like this remind me that the things worth building for have nothing to do with whatever I was annoyed about this morning. ofdollarsanddata.com/the-great-filt…

Quick take on what @lennysan and Hilary put together here - couchto5k.ai. First: - It's a great beginners guide - You could do it all over a weekend (but this was not there intention) nor am I the customer base. My only concern is if it's really necessary to be chewing through tokens with Opus 4.7. IMO - it's not. You can reap all the rewards just using Sonnet. This is a great tutorial. Highly recommend.

✌️

I just fired my entire medical team. Sold my sauna and HBOT. Cleared my shelfs of all pills. Let my blood boy go.

Been holding strategy sessions with groups of D1 athletes. Many of them clearing $250k+ in NIL this year. Built a tool that lets them see it in real time - plug in their income, toggle different strategies, and watch their tax burden drop and their capital compound. S-Corp election, SEP IRA, LLC deductions, Roth IRA. Numbers moving right in front of them. BTW - nothing is fancy about this... but this type of presentation tactic just didn't exist before. The moment it clicked for them was something. These are 19 and 20 year olds watching savings appear on a screen and realizing that money doesn't just sit - it grows for 40 years. This is what AI makes possible now. Something this polished, this personalized, built after one conversation. In 10 minutes. A tool that meets an athlete exactly where they are and shows them what their financial life could look like. That didn't exist before. It does now. Athlete first.

Late night reminder from the queen!

@krispuckett @adamnash This is incredible 😂 this is a top 5 scene for me.

I spent too much time on this: CLARK: My contention is that with Claude’s new design tool, Figma has essentially been rendered obsolete, the canvas-based paradigm is most aptly characterized as a legacy artifact from the era before models could generate production-ready interfaces from inten— WILL: [interrupting] Of course that’s your contention. You’ve never shipped anything and just watched the launch video twice. You just got finished reading some hot take, probably a Twitter thread or whoever’s got a Substack this week, and you’re gonna be convinced Figma’s dead until next month when you actually try to iterate on a flow and realize “regenerate” isn’t the same as “nudge this four pixels.” Then you’re gonna pivot to talking about how the canvas was always just a lossy interface for intent. That’s gonna last until next year when you’re in here regurgitating some take about how design tools are collapsing into a single agentic surface, you know, the post-craft utopia and the disintermediation of taste by foundation models. CLARK: [taken aback] Well as a matter of fact I won’t, because generative design drastically reduces the need for a manual canvas in the first pla— WILL: “Generative design drastically reduces the need for a manual canvas, especially as models get better at reasoning about layout and hierarchy…” You got that from that Figma-is-dead thread, right? The one that went viral last week. Yeah, I read it too. You gonna plagiarize the whole thing for us, or you have any thoughts — of your own — on this? Or is that your thing, you come into a bar, you skim some trending tweets over lunch and you pawn it off as your own idea to impress some founders, embarrass my friend? [Clark is stunned] WILL: See, the sad thing about a guy like you is in about fifty product cycles you’re gonna start doing some thinking on your own and you’re gonna come up with the fact that there are two certainties in building things. One: don’t pick a side in a tool war you’re not actually building in. And two: the people shipping right now are using Claude and Figma and Claude Code before you’ve finished writing your LinkedIn post about which one won. CLARK: Yeah, well I’ll have Claude build my whole product stack, and you’ll still be pushing rectangles around in Figma. WILL: [smiles] Yeah, maybe. But at least I’ll still know how to think when the model’s wrong.

Spent several hours this morning w/ Opus 4.7. My verdict... it feels like a regression from 4.6. I went in prepared. Detailed, refined prompts I'd built in Sonnet 4.6 specifically to save tokens + hit the ground running. The inference is faster, sure. But it reads like it's skipping the reasoning entirely. Just... output. No depth. At this cost + token burn rate, I can't justify it. The price assumes the quality is there. It's not. Not for what I'm doing. Maybe I'm missing something...? Would genuinely love to hear from people getting real value out of 4.7. I want to be wrong here. But for now, shelving it.

Ready for @AnthropicAI 's Design tool to hit desktop. Super helpful exploration today. I already use Claude as the everything tool. Assuming most people are here? I just don't here it spoken often enough. I do not open Linear/Notion/Gmail etc. I use a single tool... Claude. Now that I can mockup in Claude as well... you can see where they are taking this. I am re-designing the entire AI Support interface for our athletes. When you need to ship fast, test, iterate, + repeat. Seems like a no brainer.

Welp... still not at AGI... Opus 4.7 defaulted to... ASCII. From Opus... "that's on me. I defaulted to ASCII because I was mid-flow and it was fast, but you're right that it's the wrong output for a UX spec."

Whoever is running the GTM for @AnthropicAI is elite. It’s also nice when the product sells itself 😂 testing this today.

If everyone made the same salary, what would YOU do for work? Saw this in @JamesClear 's newsletter and immediately thought of a second-round interview years ago. The exec leaned back and asked: "If money didn't matter, what would you do?" W/ out hesitation I blurted out... teach. College professor, high school teacher, didn't matter. Something about working w/ people, watching something click, the collaboration of it all... it just came out immediately. Didn't even have to think. Then I quickly followed up w/ build. Not for equity. Not for some future outcome where I'm swimming in sheckels. But bc taking an idea that exists only in someone's head and turning it into something real people actually use? The whole product development loop is intoxicating. The messy whiteboards, the debates, the "wait, what if we…" that's the good stuff. Here's what gets me though - most ppl already know their answer. It lives right below the surface. The question just cracks it open. What does that tell you about how you're spending your Monday through Friday? BRB — calling my buddy who's still doing 80+ hour weeks in IB. Going to ask him: "Yo, would you still be fixing fonts in PowerPoint at 3am if money didn't matter?" He'll either laugh really hard, or cry. No in between. 😅 What's YOUR answer?

I love this marketing AND customer relations play. Will I buy more LMNT bc of it? 1000%. I love this product. @drinkLMNT

Not sure why this is so hard for ppl to internalize... but there is no such thing as job security. There never will be again. Snap just cut 1,000 people...their 4th major layoff in 4 years. These weren't bad employees. The business did what the business does. So here's what I'd tell any college junior right now, especially going into tech: - Don't chase the name brand. Go work your ass off on a small team in a niche space. You'll learn 10x more. You'll actually own something. And you'll build the skillset to one day build on your own. That's the only real security left. what you know how to do. apnews.com/article/snap-s…

@Slimbrady1NGMI There need to be a laugh emoji back to a response on twitter 😂

@HebardMatt Nice Akshay pick….

Recap on Claude inspired Masters fantasy pool team. 192 teams. Used Claude to build my lineup. Finished 11th. No cash. But I'm not upset about it. Only ~13% of teams rostered Rory McIlroy before the tournament. W/ out running the analysis through the model, I probably don't take him either. I went back and forth with it for a while on that call. He won by a shot. Is it luck or is it expected value actually working? The hard lesson is that good process doesn't guarantee good outcomes. You can't predict JRose bogey'ing 17. What you can do is put yourself on the right side of the probability distribution and trust it. A few things I'm taking into next year: - Treat it like science. You get better as you keep doing it. One tournament isn't a sample size. - Don't expect a great output from a single prompt. - Getting real value out of an LLM takes hours - Model selection matters. This entire process might be insanely different 12 months from now... You win some. And you learn some.