A METR study found that developers were actually slower when using AI coding tools, contrary to expectations. The research revealed that while AI excels at specific tasks like documentation, unit tests, and refactoring, it can slow developers down on complex code generation tasks. Key insights include the importance of
Table of contents
How was the METR study conducted?What was your reaction when you found out about the results?Some critics of the METR study suggested the participants just weren’t good developers. What's your take on that?Why do you think AI tools felt faster but actually weren't?You've talked about “context rot.” What is that and how should developers handle it?What’s your guidance for organizations that want to provide guidance to their developers on using AI tools?With models evolving so quickly, how do you decide when to switch or try something new?Do you think we're ready for agents to handle complex development workflows, or are we getting ahead of ourselves?Sort: