Hey there,

A few years back, a client of mine (Dave) paid me $400K a year.

Not because I wrote code for him (though my team did plenty of that). He paid me because I was the CTO in the room – the person who knew which technical decisions would cost him money six months from now, which architecture choices would scale, and which ones would become invisible tech debt that would slowly eat his business alive.

Beyond saving money, I also innovated software that multiplied the value his company delivered. So, while our professional relationship started with a $2,475 invoice, it turned into $1.9M (invoiced over several years).

As for his company?

He exited for ‘many multiples’ of that. That said, the money isn’t the point – Dave, like many non-technical founders I’ve worked with for years – knew that yes, you could pay for a ‘coder’, but the true value was having the judgment behind the code. Knowing what to build, when to ship, what to fix now vs. later, and when to say “pause…this isn’t ready.” It’s why he (and the buyers of his business) had me in the room at the exit – I spoke both the language of ‘code’ and ‘business’.

That’s part of what a CTO actually does. And for anyone that’s trying to go from idea to real software with AI as a coding assistant – decades of CTO expertise isn’t something they have (or even something that’s accessible).

But I’ll get to that. First, let me tell you why I’m thinking about all this.

The conversation too few people are having

The AI conversation right now is dominated by two extremes.

On one end, the hype crowd: “AI will replace everything, 100x your productivity, get rich quick.” If you’ve scrolled any news feed or subreddit, you know the type. Slick landing pages, social media followers, and promises that sound good until you ask a follow-up question.

On the other end, the fear crowd: “AI is dangerous, it’s taking our jobs, it’s building killer robots.” And look – some of that fear is warranted. These are real questions.

But what almost nobody is talking about is the middle.

The people who aren’t trying to build the next billion-dollar AI company. Who aren’t trying to replace human workers with bots. Who just want to use this incredible technology to build something real – something that serves people, creates value, and reflects who they are.

They’re building tools that help small businesses run better. Apps that solve problems their communities actually have. Software that makes someone’s workday easier, or their family’s life a little less chaotic.

Basically, they’re building businesses they’re proud to own. And that, I think, matters more than most people realize.

That’s the AI future I care about and it’s the one I’m building towards.

Are you building something that makes the world a little better? Something you’d be comfortable explaining to your kids? Something that, years from now, you’ll look back on and say, “I’m glad I did that”?

If the answer is yes, then you’re our kind of person.

Here’s what I actually see happening

I can describe an idea to Claude Code or Codex and watch usable software take shape in an afternoon. Things that used to take my team weeks – prototypes, tools, entire features – happen in hours now. “Vibe coding” (as folks on the Internet call it).

Fun times.

But here’s where it hits a ceiling:

Building software and building a business around software are two very different things. Same goes for making sure that software is secure enough that you’d be proud to share it – even if your target audience is just teammates or clients rather than a wide range of customers.

AI solved the access problem. Anyone can write code now. What nobody solved is the judgment problem – knowing what’s worth building, when something is ready to ship, and how to turn a prototype into something people trust enough to pay for.

That’s what I did for Dave for seven years. That’s what a CTO does. And right now, most AI-assisted founders are trying to do it without one.

One of our students, Jimmy, is an executive who’s been in business for more than 24 years and has never taken a software class. He spent two years looking for a way to learn this without learning programming languages. At one point he asked a friend who’d sold his company to Google:

There is now. You’ll meet Jimmy on Wednesday.

AI changed the starting line. It didn’t change the finish line.

You still need product-market fit. You still need users who stick around. You still need to sell. Those things were true before vibe coding and they’ll be true after whatever comes next.

The real question isn’t “can I build?” anymore.

It’s: can I build something people trust, pay for, and keep using?

Another student, Matt, asked himself exactly that. He’d paid a developer about $75,000 for the first version of his product. It worked, a customer was using it, and he needed version 2. So he went back to a dev shop with the full spec in hand, expecting a smaller number. They quoted him around $60,000.

He didn’t pay them anything. What he did instead is Thursday’s email.

What 149 IWAI readers told us

If you were one of the 149 who answered our survey about building software with AI, thank you