Interviewing candidates online is becoming difficult with AI. The risk of cheating has always been present since the candidates are not in a controlled environment, but modern AI-driven tools are making it really easy for cheaters.
Without going into cases when , it is pretty obvious that now , transforming an average guy into an expert for the duration of a video call.
Without delving into cases where candidates used deepfake technology, it is fairly obvious that AI can now assist in interviews, too, transforming an average person into an expert for the duration of a video call.
An AI can now answer all interview questions, and coding assistants can suggest what to do in a live coding session too. Assistants may be so integrated into IDEs that sometimes the candidates cannot even turn them off. Besides that, there is no way to check what’s happening on the candidate’s second screen.
Is it a problem?
Yes, and no.
The risk is to hire someone technically incompetent who happens to be good at delivering results with AI. However, that is precisely what gurus say is going to be the future: AI assistants will do the heavy lifting and humans will drive them with prompts and check the results.
The real problem, then, is whether someone can ace the interview with AI, but cannot do the job with it.
I see two scenarios:
1 - the incompetent candidates get the job and fail to be good at it. They get fired during their probation period, and everyone would have lost a lot of time.
2 - The incompetent candidates can let the AI do the job, and build their career with that, while knowledgeable dinosaurs will become unable to keep up.
I hope for the former, but it’s going to be the latter.