“A week after TechCrunch profiled Anthropic’s experiment to task the company’s Claude AI models with writing blog posts, Anthropic wound down the blog and redirected the address to its homepage.