Google Pilots Gemini in Engineer Interviews Focusing on Prompting and Validation
AI assistants · AI coding
Google is piloting the use of its AI assistant Gemini in select software engineering interviews, shifting focus toward evaluating prompt engineering, output validation, and debugging. Recruitment VP Brian Ong noted the effort aims to mirror actual team workflows, while a spokesperson stressed the initiative remains an experimental pilot.
According to internal documents reported by Business Insider, Google is adjusting its engineering recruitment process to evaluate candidates' ability to work with AI tools. During 'code comprehension' interviews, candidates are permitted to use Google-approved Gemini to read, optimize, and debug existing code. Interviewers assess proficiency in prompt engineering, output validation, and verifying underlying program logic. Brian Ong, Google's VP of Recruiting, stated the pilot aims to align hiring with daily workflows; a company spokesperson noted the format is an ongoing trial and not yet deployed across all interviews.
Key takeaways
- Google is piloting Gemini usage in select code comprehension interviews to assess prompt engineering and output validation skills.
- Recruiting VP Brian Ong stated the pilot mirrors real workflows, while a spokesperson emphasized it remains an experimental trial.
- Google reported earlier in April that 75% of its new code was AI-generated before human review, prompting changes in technical evaluation.
Source：數位時代 Read the original ↗
Prepared, checked and automatically published by AI; not individually reviewed by a human. Translation checked by AI.