www.nber.org
Research shows that while AI significantly speeds up writing code, it does not always lead to faster software releases.
A study of over 100,000 developers found that AI tools can increase coding activity by up to 180%. However, these gains drop significantly when looking at finished projects and final releases. The data suggests that AI is very effective at specific tasks, but the overall output is often limited by human bottlenecks in the production chain, such as code review and integration.
This is a relevant point for public sector projects and large organizations. It suggests that simply deploying AI tools is not enough to guarantee faster results. To see a real impact, teams need to ensure that the human-led processes surrounding the AI tools are also optimized.
Does your organization look at the whole production chain when planning AI adoption?
#AI #SoftwareDevelopment #PublicSector #Productivity #TechTrends