Had my exit interview with the new (started 4 days ago) CTO yesterday. Things that stood out:
1. He was surprised to learn that lines of test code should outnumber lines of actual code.
2. He suggested that if he defines a spec and gives that to a code generator, then it would be useless to have the same code generator generate all these unit tests.
3. He couldn’t understand why the number of tests would explode if he intended to thoroughly test the application at the highest level only, ie the output of his code generator.
4. Concepts he looked up during our chat: Testing Pyramid, DDD, Clean Architecture.
5. He studiously avoided saying AI, talking only about a mythical code generator that would unerringly translate his presumably complete spec to code.
Anyway, is it time yet to start a consultancy to help companies fix the problems caused by the parrot?