Comment by _whiteCaps_
20 hours ago
A surprise live test is absolutely the wrong approach for validating whether someone's done the work. IMO the correct approach is to go through the existing code with the applicant and have them explain how it works. Someone who used AI to build it (or in the past had someone else build it for them) wouldn't be able to do a deep dive into the code.
We did go into the assignment after I gently bowed out of the goofy live test. The CTO seemed uninterested & unfamiliar with it after returning from a 3 week vacation during the whole process. I waited. Was happy to run him through it all. Talked about how to extend this to a real-world scenario and all that, which I did fantastically well at.
I feel your pain. This isn't a question about AI or not. It's about if you can do the work and do it well. This kind of nonsense happened before AI. If you can't win the game of Jeapordy you don't get the job which has nothing to do with being a Jeapordy contestant!