An Anthropic engineering manager has dropped the coding interview from his team’s hiring process, saying Claude now solves the puzzles better than most applicants.
Felix Rieseberg, an engineering lead on Claude.ai and the Claude Code desktop app, said on X that the coding interview has stopped telling him anything useful. He removed the round from his team’s hiring loops this month.
Rieseberg wrote in a longer blog post that watching candidates define a problem and reach an elegant solution used to be a weak but real predictor of how they would work on his teams. After Claude became good at these exercises, he said, a small React challenge told him about as little as an assembly language quiz.
Anthropic’s performance team ran into the same problem with its take-home test. Team lead Tristan Hume wrote in January that Claude Opus 4 beat most applicants within the time limit, so he made the test harder. Then Opus 4.5 came out and kept pace with the best candidates. The current version leans on odd, puzzle-game-style problems that Opus 4.5 couldn’t crack.
Rieseberg has not yet found a replacement for the coding round. In the meantime, he is going through resumes and portfolios, looking for small prototypes that show taste, empathy, and principled thinking. He also asked other managers to share any better methods they had come up with. Jaana Dogan, a principal engineer at Google, replied on X that she would rather watch candidates work with coding agents, since that tells her more about how well they code.
Rieseberg’s essay is mostly for junior engineers. Its title, “Hunger beats tenure,” is the argument: building is so cheap now that newer engineers can beat veterans. His evidence is two fairly new Anthropic engineers who built a working graphical interface for Claude Code in one weekend. The first public version shipped about eight days later.
The wider talk about AI and new graduates is gloomier. The hiring data is mixed. ZipRecruiter labor economist Nicole Bachaud told The Register that tech employers still test for skills most often. In the company’s October survey of 1,500 employers, 62.5 percent of those in the information sector required formal skills tests at the application stage. About 70 percent now proctor those tests or hold them in person to keep AI out.
Anthropic still asks candidates not to use Claude in live interviews or take-home tests, unless told otherwise. Rieseberg dropped the coding round for his team only. Applicants elsewhere at the company will likely still face one.
Maybe you would like other interesting articles?