A longitudinal study analyzing data from 40 companies found that while AI usage in engineering teams increased by 65%, productivity measured by pull request throughput only rose by about 10%, far below the 2-3x gains often promised. The main reason is that coding is just one part of engineers’ work, with significant time still spent on planning, alignment, and reviews, which AI has not yet significantly improved.
This is an ainewsarticles.com news flash; the original news article can be found here: Read the Full Article…
