Anthropic's AI Fluency Index finds that polished AI output makes users less likely to check for errors

Anthropic analyzed nearly 10,000 Claude conversations and found that users get less critical when AI output looks polished. The company's new AI Fluency Index also shows that iteration is the strongest predictor of competent AI use, but comes with its own tradeoffs. The article Anthropic's AI Fluenc