Code is not just a commodity - it is a liability, at least production code. You build it - you own it.
So as soon as PMs can do maintenance of the code that AI wrote for them in production, I’m fine with that. But if it remains the sole responsibility of engineers, then engineering will again become the bottleneck and the holders of “exclusivity”.
That said, it’s great when PMs use AI for POCs, demos and prototypes - it explains ideas better than any documentation. But using such solutions as-is in production sounds like a nightmare for those who would need to maintain them over the years.
I think code becomes a commodity in this transformation, as assembler code became one tens of years ago. The more I work with AI, the less I care about how it's implemented under the hood, as long as it meets my constraints and evals (from code architecture to testing principles, etc.).
If it meets my performance expectations, passes tests, keeps SLOs, and has pleasing UI/UX, why should I bother with what's under the hood? 🤔
Code is not just a commodity - it is a liability, at least production code. You build it - you own it.
So as soon as PMs can do maintenance of the code that AI wrote for them in production, I’m fine with that. But if it remains the sole responsibility of engineers, then engineering will again become the bottleneck and the holders of “exclusivity”.
That said, it’s great when PMs use AI for POCs, demos and prototypes - it explains ideas better than any documentation. But using such solutions as-is in production sounds like a nightmare for those who would need to maintain them over the years.
I think code becomes a commodity in this transformation, as assembler code became one tens of years ago. The more I work with AI, the less I care about how it's implemented under the hood, as long as it meets my constraints and evals (from code architecture to testing principles, etc.).
If it meets my performance expectations, passes tests, keeps SLOs, and has pleasing UI/UX, why should I bother with what's under the hood? 🤔
This isn’t about losing magic, it’s about learning a new kind.
💯
Not an easy thing, but our roles as engineers must evolve too.