The experiment conducted by interviewing.io showed that using ChatGPT in technical interviews went undetected by interviewers. The post argues that biasing hiring towards candidates who over-prepare for standardized problems is a greater risk than cheating. It also suggests that the ability to critically analyze and make adjustments to code generated by LLMs is a valuable skill for developers. The post advises against trying to eliminate the use of LLMs and instead recommends creating a humane hiring process.
Sort: