Founders Are Building Again

I was having a beer with a founder the other week. He runs a nine-figure company. Proper grown-up business, hundreds of employees, the works. He turned up late (which is fine, founders are always late) but the reason was unusual. He texted me: "Sorry, have to wait for my agents to finish."

He arrived at the pub with his laptop still open. Every few minutes he'd glance down at it, check something, nod to himself, and carry on chatting. When I asked what he was working on, he lit up. He'd been shipping features. Actual features. In his own product. A CEO who hadn't written code in years was now personally building things with AI tools, and he couldn't stop talking about it.

I told him I'd done the same thing while waiting for him. Kicked off a couple of changes on Claude Code Web from my phone. We looked at each other and laughed. Two blokes at a pub, both building software between sips of beer. It felt like 2009 again.

This Isn't a Trend. It's a Correction.

I've been in this game for 20 years now. I spent a decade at Google, built startups, ran engineering teams, hosted The Startup Podcast. I've seen a lot of "paradigm shifts" that turned out to be incremental improvements wearing a good marketing hat.

This does feel different.

What I'm seeing (not theorising about, but actually witnessing in real conversations with real founders) is that senior technical leaders are spending more time building than they have in years. CTOs who graduated to full-time management three roles ago are suddenly prototyping again. CEOs are shipping bug fixes on weekends. Not because they have to. Because they want to.

The friction that pushed founders out of the codebase (the setup, the boilerplate, the context-switching tax of dipping in and out) has largely evaporated. When you can describe what you want and have an AI agent build it while you're at the pub, the calculus changes entirely.

The Joy of Building (Redux)

Most founders miss building. Nobody talks about this. They didn't stop because they lost interest. They stopped because the opportunity cost became too high. When your company hits a certain scale, spending four hours debugging a webpack config is irresponsible. So you delegate. You hire. You manage. And slowly, the thing that got you into this industry in the first place, the mesmerising act of making something from nothing, fades into memory.

AI tools have given that back.

I spoke to one founder who told me he'd shipped more code in the last six months than in the previous ten years combined. Another said it had changed how he hires dev teams and runs product. Completely. These aren't wide-eyed early adopters chasing a hype cycle. These are battle-scarred operators who've seen everything, and they're excited again.

That excitement matters more than people realise. A founder who is personally engaged with the product — who understands the codebase, who feels the friction points first-hand — makes better decisions. Full stop. The best companies I've seen in 2026 are the ones where the founder hasn't just approved the roadmap. They've built half of it.

What This Means for Everyone Else

If you're an engineer at a company where the CEO is vibe coding on weekends, that's a signal. A good one. It means leadership actually understands what your tools can do. It means when you ask for unlimited AI tokens, you're not explaining the concept to someone who's never used them. It means the organisation is going to move faster, because the person at the top has felt the speed themselves.

Conversely, if your leadership still thinks AI coding tools are "a nice experiment" or "something for junior devs to play with," well, that tells you something too. The gap between companies that get this and companies that don't is widening every week. And it's widening from the top down.

This is why we built the directory. Because the companies where founders are building again, where AI tools aren't just permitted but actively championed by leadership. Those are the places you want to be. The culture is different. The pace is different. The relationship between humans and tools is different.

The Uncomfortable Truth

I'll be honest: this shift isn't comfortable for everyone. When a CEO can prototype a feature in an afternoon that would have taken a team a sprint, it raises awkward questions about team size, velocity expectations, and what "senior" even means anymore. I've had those conversations too. Founders quietly wondering whether they need as many engineers as they thought.

But that's the wrong framing. (It usually is.)

The right framing is: what happens when everyone in your organisation can build at ten times the speed? When the bottleneck isn't engineering capacity but imagination and taste? The founders who are building again aren't replacing their teams. They're raising the floor. They're compressing the feedback loop between "I think we should try X" and "here, I built a prototype of X, what do you reckon?"

That loop — idea to prototype in hours instead of weeks — changes everything about how a startup operates.

Building Is Believing

I keep coming back to that evening at the pub. Two people running companies, both with laptops open, both building things. Not managing. Not strategising. Not "aligning stakeholders." Building.

After 20 years in tech, I can tell you: the energy right now reminds me of the early days of the internet. That feeling that the rules have changed and nobody's quite figured out the new ones yet. Except this time, the founders aren't just funding the revolution. They're writing the code.

If you're a founder who hasn't opened a terminal in years, try it. Seriously. Spin up Claude Code, describe what you want, and watch what happens. You might surprise yourself.

And if you're an engineer looking for a company where leadership understands AI tools, where the founder builds alongside you, not just above you: browse the directory. These companies exist. And they're hiring.