<ul data-eligibleForWebStory="false"><li>AI coding tools like Cursor and GitHub Copilot have promised to enhance developer productivity by automatically writing code, fixing bugs, and testing changes using AI models from various companies.</li><li>A study by non-profit AI research group METR raises doubts about the productivity gains of AI coding tools. In a trial with experienced developers, using AI tools actually increased completion time by 19% instead of reducing it by the forecasted 24%.</li><li>The study indicates that developers spend more time prompting AI and waiting for responses when using these tools, leading to slower workflows. AI also struggles in large and complex code bases, further hindering its efficacy in this context.</li><li>While the study doesn't definitively conclude that AI systems fail to speed up most developers, it highlights the potential challenges and limitations of AI coding tools in certain scenarios, cautioning against assuming universal productivity gains.</li></ul>

AI coding tools may not speed up every developer, study shows

Discover more