In the post shared on X, Cherny revealed that he was surprised that people were surprised by the way he prompts Claude. Anthropic’s Claude Code creator, Boris Cherny has offered a simple piece of advice to users who are trying to get better results from AI models. In a post shared on social media platform X (formerly known as Twitter) Cherny asked the users to talk to Claude the same way they would talk to a coworker. As per Cherny users no longer need highly structured and very detailed prompts for most tasks, as modern AI models are increasingly capable of figuring out how to achieve a goal on their own. “Talk to Claude the way you would a coworker. There’s no secret to prompting. There’s no need to be overly scaffolded or prescriptive for most tasks. Give Claude a goal, and it will figure it out,” he wrote.
Cherny contrasted the current approach with the early days of Claude’s Sonnet 3.5 models, when the exact wording of a prompt mattered a great deal.
Today’s models, he suggested, need less hand-holding on the “how” and more clarity about the “what. The prompt asked Claude to make the output “beautiful and interactive,” focus on business wisdom from the episode, use extensive reasoning and iteration, and incorporate elements such as watercolour illustrations, historical photographs and visual storytelling inspired by publications such as The New Yorker and The New York Times. Alongside his comments, Cherny shared an example of a prompt he uses with Claude.