For the past few years, most of us have interacted with AI in a similar way. We’ve used it to write, brainstorm, summarize, and clean up ideas. It’s been helpful, sometimes surprising, and occasionally impressive.
But there’s a big difference between using AI as an assistant and realizing you can actually build with it. That realization is what led me to start Just Build It, a short podcast and video series about what happens when AI collapses the distance between idea and implementation.
The first episode features my friend James Raybould, and one thing I appreciated about our conversation is how grounded his perspective was. James had already been a heavy AI user for a while, but his real turning point came when he realized he had become one of those executives telling everyone else to build prototypes without actually using the tools himself.
So he jumped in.
He started with something simple (a workout tracker) and quickly had the experience that seems to flip the switch for a lot of people. Something that might have taken a team of engineers a week or two to build was now good enough in an hour or two.
That’s when it stopped feeling theoretical.
One thing James emphasized is that building with AI isn’t magic. Things break. APIs are confusing. Sometimes the model goes in circles. His framing was that AI today is like the most brilliant intern imaginable — but still an intern. You still have to guide it, correct it, and occasionally tell it very bluntly that it’s wrong.
But the feedback loop between idea and prototype is dramatically shorter than it used to be. Once James started building regularly, he also discovered how quickly things compound. Once you have a basic stack and workflow, each new project becomes easier than the last. And the things you build don’t have to be massive startup ideas.
In one example he shared, his daughter complained that her outfit didn’t match on the way to school. Over the weekend he built a wardrobe picker inspired by Clueless. In another, he built a simple breathing app using his own voice to help his kids calm down.
None of these ideas required a company roadmap or engineering team. They just required someone close to a problem. That’s the shift.
For most of the history of software, the hardest part was building something at all. Now the harder question may simply be: What’s actually worth building?
That’s the question Just Build It is exploring. The first episode with James Raybould is live now.
Full episode here: