Rather than banning AI in technical interviews, better to see how the candidates use it and if they can comprehend what the LLM is saying, the quality of their prompts and own thinking.
Rather than banning AI in technical interviews, better to see how the candidates use it and if they can comprehend what the LLM is saying, the quality of their prompts and own thinking.
I believe people who are using these AI tools to pass interviews wont be able to use AI in their real job in a net positive manner.
I recently hired two engineers that were good at clearing the interview rounds using AI -- I knew because I encouraged them to use AI.
But when it came to large complex codebase or problems that required critical thinking everything fell apart.
I couldn't agree more. LLMs are legitimate tools and, ideally, I want to see how effective a candidate is in using their available tools to solve complex problems.
The service on offer here is different. It's providing a means to use LLMs to fake your way through a technical interview.
Showing that you can use LLMs to quickly and correctly solve problems is a good skill to have. Offering up a solution from an LLM as your own work without acknowledging how you got there is just misrepresentation... or to put it another way is just lying. Maybe fake your degrees and experience while you're at it, right?
At least in the long run, many that need these tools to get in will be found out once they start having to solve real problems on the job. Just a shame about other, more qualified people being turned away. Of course if the LLM was sufficient enough on its own, perhaps a real software developer was never required to begin with.