Ensuring code works in the wild
After you write code, there is always a possibility that you will run into bugs. It is a stressful for the development team and the customer when that happens, since the customer can't get work done, and the team can panic getting a patch out. There are tests you can write, but it's difficult to make comprehensive tests with good coverage. Using penetration testing is similar. Plus, if you re-write or re-implement the code, the tests can get outdated, which means you must rewrite it.
Discussion (2)
Maybe we can have a AI test the code. Such as an AI Pen tester that generates fuzzing on inputs to try to break the sanitizers.
It can also rely on generating its own tooling (i.e. Python functions) so that it doesn't waste LLM calls to create 100 A's to cause overflow.
What happens next?
Fires with the most votes get noticed by builders and entrepreneurs looking for problems worth solving. Share this fire with others who experience the same problem to increase its visibility.