Last updated on January 26, 2026 by Editorial Team
Author(s): Ahmed M. Abdelfattah
Originally published on Towards AI.
16 developers. 246 actual work. A randomized controlled trial. The results contradict every claim of the sellers. And everything the developers believe.
Same study. Same developers. Same work. Three numbers that should not exist together.
The article discusses a study conducted by METR, which found that developers using AI tools believed they were faster, yet measured results showed a 19% slowdown in productivity. Initial expectations were a 24% increase in speeds, as opposed to the reported 20% increase, highlighting a significant difference in perception versus reality – 39 percentage points. The study results suggest that developers may feel more productive due to the enjoyment associated with AI tools rather than the actual output, which may lead to a misleading assessment of the effectiveness of AI in software development.
Read the entire blog for free on Medium.
Published via Towards AI
Take our 90+ lessons from Beginner to Advanced LLM Developer Certification: This is the most comprehensive and practical LLM course, from choosing a project to deploying a working product!
Towards AI has published Building LLM for Production – our 470+ page guide to mastering the LLM with practical projects and expert insights!
Find your dream AI career at Towards AI Jobs
Towards AI has created a job board specifically tailored to machine learning and data science jobs and skills. Our software searches for live AI jobs every hour, labels and categorizes them and makes them easily searchable. Search over 40,000 live jobs on AI Jobs today!
Comment: The content represents the views of the contributing authors and not those of AI.
