Preliminary data from a longitudinal AI impact study
newsletter.getdx.com
March 11, 2026
2 min read
Summary
Preliminary data from a longitudinal AI impact study indicates that AI productivity gains are approximately 10%, rather than the previously suggested 10x. High expectations for AI productivity, influenced by social media and vendor marketing, may not align with these findings.
Key Takeaways
- AI productivity gains in engineering are approximately 10%, significantly lower than the 2-3x increases often suggested by marketing.
- During a study of 40 companies, AI usage increased by an average of 65%, but pull request throughput only rose by 9.97%.
- Developers report that coding is not the primary bottleneck in their workflow, as planning and other human-centric tasks remain largely unchanged.
- The ongoing study will further investigate why some teams achieve greater productivity gains with AI and how leaders can facilitate this.
Community Sentiment
MixedPositives
- A broad 10% productivity improvement across industry is significant, indicating that AI's impact could be transformative, even if it doesn't seem headline-worthy.
- LLMs have effectively reduced the cost of deployment, allowing for quicker iterations and potentially reshaping how coding is approached in the future.
- The rapid iteration capabilities of AI tools are enabling programming teams to work much more efficiently, suggesting a shift in productivity dynamics.
Concerns
- The marginal effect on PR throughput raises concerns about the actual productivity gains from AI, suggesting that the improvements may not be as impactful as anticipated.
- Relying on PR throughput as a metric for AI productivity is flawed, as it does not capture the true potential of AI in transforming development processes.
- Skepticism around the validity of AI productivity enhancements from 2024 data highlights the uncertainty and potential overestimation of AI's impact on developer roles.
Related Articles
Productivity gains from AI coding assistants haven’t budged past 10% – survey
Feb 19, 2026
AI doesn’t reduce work, it intensifies it
Feb 10, 2026
Coding assistants are solving the wrong problem
Feb 3, 2026
Why developers using AI are working longer hours
Mar 7, 2026
Will Claude Code ruin our team?
Mar 8, 2026
Source
newsletter.getdx.com
Published
March 11, 2026
Reading Time
2 minutes
Relevance Score
45/100
Why It Matters
This page is optimized for focused reading: quick context up top, a clean summary block, and a direct path to the original source when you want the full story.