AI coding agents generate more code, but not more software

A recent study has found that AI-powered coding agents can generate more code than human developers, but this increased productivity is not necessarily translating to more software being developed. This is because the efficiency gains from AI-generated code are often absorbed by the human review process, which remains a bottleneck in software development. The study suggests that while AI can automate certain tasks, it cannot replace the complex decision-making and review processes that come with human development.