Generative AI, including large language models (LLMs), can produce significant amounts of code. However, measuring productivity solely by lines of code generated is considered an inadequate metric for assessing software development output.

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.

AI's coding capabilities are advancing faster than effective utilization by engineering teams. Successful product development, such as Anthropic's Cowork, illustrates the importance of bridging the gap between AI capabilities and practical application.

A survey of nearly 5,000 technology professionals revealed that 90 percent are using AI at work, with over 80 percent reporting increased productivity. Despite these benefits, developers are reportedly working longer hours.

METR's previous study indicated a 20% slowdown in task completion among experienced open-source developers using AI tools. A new experiment launched in August 2025 aims to assess AI's impact on developer productivity over time, but participant feedback suggests the data may be unreliable.