AI made building fast. Now engineers have to think like product people.
Hi there,
Thanks for signing up. This is the first issue, so let me say what you just got yourself into.
I've been writing software since I was 14. I went from engineer to CTO to Chief Product Officer at GamerSafer, and at night and on weekends I build tools for people who make software with AI: shpyrd puts your app online, Seed lets an app grow itself, and Percy is sending you this email. This newsletter is about building products in this new era, and about the people who build them.
The distance between an idea and a product just collapsed
Building software has always been hard. For most of my career, the expensive part of a new idea was building it. You needed a team, weeks of work, and a lot of meetings before anyone outside could touch it. Most ideas died in that gap, not because they were bad, but because they never got tested.
That has changed. With AI, one person can go from an idea to something real people can use in days, sometimes in an afternoon. Building is no longer the bottleneck.
But here's the thing: building was never the hardest part. The hardest part was always getting customers. Finding the people who have the problem, convincing them to try something new, and earning enough trust that they keep using it and pay for it. AI made building fast. It didn't solve that other part.
So the bottleneck moves to the questions that were always the hard ones: what should we build, for whom, how do we reach them, and how will we know it worked?
Engineers were taught to stay away from product
A lot of engineers learned, explicitly or not, that product is someone else's job. Take the ticket, write the code, close the ticket. Talking to customers, questioning the requirement, caring about the metric: that was for product managers.
I think that was always a mistake, and AI makes it an expensive one. When writing code gets cheap, an engineer who only writes code is competing with the cheapest part of the process. An engineer who understands the user, the problem and the business is suddenly the most valuable person in the room, because they can take an idea all the way to customers on their own.
What getting closer to product looks like in practice:
- Ask why before how. Know the problem a feature solves before you open the editor.
- Talk to the people who use what you build. Even one conversation changes what you write.
- Ship small and early, then watch what actually happens. Real usage beats any spec.
- Care about the outcome, not the merged pull request. The work is done when it helps someone.
None of this means engineering matters less. Software still has to be safe, and it still has to keep running. It means the best engineers of the next few years will be the ones who can hold both: the craft and the customer.
That is what I'll write about here: building products with AI, getting ideas to real people faster, and helping engineers grow into people who own outcomes.
That's it for the first one. Reply and tell me: were you taught to stay away from product too? I read every email.
Marcelo