I spent 3 months "vibe coding" with Lovable as a non-technical PM.
It's both magical and infuriating.
My conclusion is that non-technical people CAN build simple prototypes, websites and internal tools, but would struggle to build production-grade products without any technical expertise. Think of current AI as a junior dev with outstanding syntax knowledge but terrible judgment.
Here are some things I learned in the last 3 months that seem to work well:
1. Treat it like a software development intern (write PRDs, user stories, acceptance criteria)
2. Work in tiny increments—big changes confuse the hell out of it
3. Use the "Uncle Bob" persona for cleaner architecture
4. Always refactor when prompted (but test before/after)
5. Don't be afraid to revert and start over—code is now the cheap part
Insight about the future of product: When writing code becomes cheap, many experiments that would've been discarded can now be released. We're transitioning from discovery-heavy processes to rapid iteration in production. Engineers won't vanish—they'll take on more PM responsibilities. And PMs/UXers need to learn some engineering to ship independently too.
The best advice I can give you: if you're a product person still sitting on the sidelines, what are you waiting for? Start building.
I am a product leader with 12 years experience in consumer facing products, both B2C and B2B. The ideal company for me is ambitious, data driven, >100 employees and growing or almost entering the growth stage.
Atlas Shrugged is indeed mind blowing. The reason is because you can actually see this sort of thing happening (depending on where you are, in my case, Brazil).
I wrote the article on Medium. I was in contact with 2-3 ex-students of that course. None of them received refunds several months after my article (I imagine I received the refund because I started the "movement").
I have failed (about 50k people saw how I failed and what I learned here: http://www.sergioschuler.com/startup-lessons-learned-from-my...) and now I work for a pretty normal company, as their digital marketing head (not as impressive as it sounds, but I enjoy it).
I am building something very simple on the side (libertarian tshirts, since the idea is growing here in Brazil) and I still crave to solve problems.
Your evaluation certainly have merits and a lot of it we confirmed when speaking to clients later. Just one thing to make clear: I have 5+ years of team management/leadership development and I used a similar tool I created and evolved (in paper form) to improve a loooot of teams. So the tool works - BUT you are correct, it is a pain to apply and you probably need consultancy to get out of the team management hole if you are not a skilled manager.
It's both magical and infuriating.
My conclusion is that non-technical people CAN build simple prototypes, websites and internal tools, but would struggle to build production-grade products without any technical expertise. Think of current AI as a junior dev with outstanding syntax knowledge but terrible judgment.
Here are some things I learned in the last 3 months that seem to work well:
1. Treat it like a software development intern (write PRDs, user stories, acceptance criteria)
2. Work in tiny increments—big changes confuse the hell out of it
3. Use the "Uncle Bob" persona for cleaner architecture
4. Always refactor when prompted (but test before/after)
5. Don't be afraid to revert and start over—code is now the cheap part
Insight about the future of product: When writing code becomes cheap, many experiments that would've been discarded can now be released. We're transitioning from discovery-heavy processes to rapid iteration in production. Engineers won't vanish—they'll take on more PM responsibilities. And PMs/UXers need to learn some engineering to ship independently too.
The best advice I can give you: if you're a product person still sitting on the sidelines, what are you waiting for? Start building.