Keep pulling the thread on Claude Code.
According to McKinsey research, 95% of its enterprise clients report not seeing a P&L impact from their AI initiatives.
In October 2025, over 40% of enterprise AI initiatives were canceled, a 2.5-fold increase from the previous year.
Only 1% of enterprises report having mature AI deployments that deliver significant value.
A benchmark study by METR in May 2025 found that developer productivity declined by approximately 20% when using early AI coding tools, despite developers expecting a 30% gain.
A follow-up study in January 2026 showed that developer productivity with AI tools improved to an absolute gain of 20%.
The volume of AI-generated code committed to GitHub via Claude Code increased sevenfold since October 2025.
The release of models like Gemini 3 and Claude Opus 4.5 in October 2025 was the single factor that drove the significant improvement in developer productivity.
The Claude 4.6 Opus model can automate nearly 15 hours of developer work with a single prompt.
Data center investments for AI are projected to reach approximately $7 to $8 trillion by the end of 2030.
To realize the value of projected AI investments, enterprises would need to automate 15% to 25% of their white-collar work.
By the end of 2026, LLMs are expected to be able to access and use enterprise tools and PCs at a level comparable to human employees.
AI has the potential to automate 44% of all jobs in the US economy, which represents about 68% of all white-collar jobs.