Key facts
- Mercor is implementing traditional whiteboarding interviews, a strategy that Google and Microsoft are reportedly moving away from.
- The company aims to assess candidates' understanding of AI tools, statistics, judgment, and systems design through these sessions.
- Mercor's approach contrasts with the trend of allowing AI coding assistants and take-home assignments in technical interviews.
- Osvald Nitski, Mercor's head of product, emphasized the importance of human judgment over simply regurgitating AI outputs.
Mercor, an AI training company, is bucking the trend in Silicon Valley by reverting to traditional whiteboarding interviews instead of embracing AI-assisted coding assessments. Osvald Nitski, Mercor's head of product, explained on the "20VC" podcast that the company wants to ensure candidates possess strong judgment and understanding of statistics and systems design, rather than relying on AI tools like Claude to generate answers.
While Mercor utilizes AI for screening candidates in its talent marketplace, its full-time hiring process deliberately avoids take-home assignments and work trials that have become popular alternatives to standard interviews. Nitski expressed concern that AI makes it too easy for individuals to "offload" decision-making, potentially hindering their ability to think critically and independently.
This approach stands in contrast to major tech firms such as Google, Microsoft-owned LinkedIn, and Cisco, which are reportedly adapting their interview processes to permit the use of AI tools. Startups like Lovable, Cursor, and Kilo are also favoring work experiments, and companies like Cognition are actively encouraging candidates to use AI coding assistants during interviews, likening it to using a calculator for a math test.