A deep understanding of the problem space is necessary for building good products.

I've worked on teams where delivering was more important than whether the deliverables actually made any sense, because stakeholders valued "seeing" progress.

When no one steps up to absorb that pressure and manage expectations and timelines, you end up with poorly built products that are effectively useless.

A good-enough product is often not good enough. Good work takes time, even with AI-assisted tooling, because you simply can't offload domain knowledge.

Here's my hot take: as a builder, if you can't clearly articulate the problem space with all its nuances and explain what you've built, the product is probably bad. Adoption and revenue can certainly prove that the product is viable, but they don't prove the team understood the problem.

If I take this too far: we're headed toward an end where no one knows anything and consciousness, pure existence, becomes the ultimate abstraction.