According to @AnthropicAI, the new AI Fluency Index tracks 11 user behaviors across thousands of Claude.ai conversations to quantify how effectively people collaborate with AI, such as iteration frequency and refinement patterns; as reported by Anthropic’s research page, the index aims to help teams benchmark prompt practices and training impact, while @emollick noted the framework is debatable but praised its practical guidance. According to Anthropic, the dataset highlights behaviors correlated with higher-quality outcomes, offering enterprises a measurement framework to design upskilling programs, evaluate prompt engineering playbooks, and set KPIs for AI-assisted workflows.