Howdy Founders,

No new podcast episode today. We don’t have a nanny at home right now and 2.5 year olds are actually quite a lot of work. Since I’m perpetually unemployed (unemployable?), I will father the child. I will teach him the “way of the concrete” (I own a concrete business now. More on that next week).

Anyway, here are three things I really hope you’ll take a second to think about this week or maybe over the weekend. As always, let me know if I hit the mark or if you disagree.

The founder who couldn't stop — and what it cost her

Rachel Draelos is a doctor and a PhD. She spent seven years building Cydoc — an AI-native EHR that automated patient history intake. She had patents. She had paying customers. She had press. She had grants. By every metric that startup advice articles tell you to track, she was making progress.

She still shut it down. Revenue never broke $28K a year.

But here's the part that stuck with me. It wasn't the product that failed. It wasn't the market. What kept her going — and going, and going — was her identity. She was the kind of person who doesn't quit. She'd staked a decade of her life on this, passed on residency, coded from a hospital bed at one point. Quitting wasn't just a business decision. It was a statement about who she was.

And that's the trap.

The optimism bias that carried her through every hard stretch — "one more pivot, one more customer, one more adjustment" — is the same trait that makes founders worth betting on in the first place. Resilience and denial look identical from the inside. That's what makes this so hard to write about and so easy to fall into.

She's written the full postmortem, and it's one of the more honest things a founder has published in a while. I recommend checking it out.

The question worth sitting with this week:
Are you still building because the path forward is real, or because stopping would feel like a betrayal of who you are?

Those are different questions. One is about the business. The other is about you.

The permission structure just dropped — and I've been waiting four years for people to take this seriously

Jack Dorsey fired 40% of Block's workforce. Four thousand people. He said the reason was AI — not restructuring, not performance — AI. The stock went up 18%.

Then he went on a media tour and said: most companies will do the same within a year.

I've been saying a version of this for four years. Not to be right about it. But because I think the founders and people in your life who haven't started thinking about this are going to get caught flat-footed, and the window to get ahead of it is not as wide as it looks.

The sharpest line I read in the coverage came from a UVA Darden analysis: "Musk then, Dorsey now, just provided tech leaders with the permission structure to finally right-size." That's what actually happened here. It's not that AI is new. It's that a major CEO did it publicly, said the quiet part loud, and got rewarded for it. Every CFO in America screenshotted that stock chart. The conversations happening in boardrooms this week are different than they were a month ago.

There's a version of this that's scary, and there's a version of this that's a massive opportunity — depending on where you're standing.

I talked about this recently with Jeff Price on the podcast — I'll link that below — and he made a case I keep coming back to: boring businesses might be the future, not the past. Not every founder needs to be building a rocketship. Some of the most durable income in the next decade will probably come from things that sound unglamorous, that AI can't easily replace with a prompt, that have been hiding in plain sight as "old economy" businesses. Think about giving it a listen HERE.

The future-proofing conversation isn't new. It just got a lot louder.

OpenAI killed Sora — and it's the most important product lesson of the year

OpenAI shut down Sora last week.

If you weren't tracking it closely: Sora was their text-to-video product. It launched six months ago to massive fanfare. Waitlists. Celebrities posting clips. "The future of filmmaking" take-pieces. And then: roughly $15 million per day in inference costs, user retention under 8% after 30 days, and a shutdown announcement that Disney found out about one hour before the general public. Disney had committed a billion dollars to a collaboration. Past tense now.

WIRED framed it cleanly: Sam Altman had been running OpenAI like a YC portfolio. Bottom-up culture, teams chasing whatever was interesting, a dozen things in flight at once. That culture produced extraordinary research output. It apparently also produced pretty terrible product discipline.

This is the most expensive version of a lesson most founders learn at much lower cost: building something impressive and building something that works as a business are not the same problem. Sora worked. It generated genuinely stunning video. It still got killed because the economics never closed.

The uncomfortable question for founders building in AI right now — especially anything compute-heavy — is: are you building toward a product, or toward a demo? The gap between "this is amazing" and "this is a product people will pay for at a margin that works" is wider than it looks from the demo stage. (I SHOULD KNOW). Sora had OpenAI's brand, OpenAI's compute, and the most buzzed-about launch of 2025. It still couldn't close that gap.

The other lesson is about focus. Altman is apparently now entering his "focus era" which means: fewer products, narrower portfolio, preparing for an IPO. When even OpenAI has to learn this the hard way, it might be worth asking yourself how many things you're currently building that you couldn't kill if you had to.

THE END.

That’s it for this week as my child’s naptime has been over for like 15 minutes. Gotta run!

Love you!