Felix Rieseberg, an engineering manager at Anthropic, no longer uses traditional coding interviews when hiring for his team. He argues that AI models such as Claude now outperform humans at some coding tasks, reducing the ability of these tests to distinguish strong candidates. He wants to prioritize ambition, curiosity, taste, empathy and the ability to ship projects quickly with AI tools. Portfolios and self-initiated projects are under consideration, but no complete replacement method has yet been implemented. The approach also creates evaluation and fairness challenges, since personal projects favor candidates with free time.
Source: Read the original article


