Anthropic co founder Boris Cherny unveils loop engineering concept
Boris Cherny, co-founder of Anthropic and creator of Claude Code, says AI is moving past manual prompts.
In a recent interview, he introduced "loop engineering," where AI agents now generate and refine their own instructions, so humans barely need to step in.
Agents write prompts, may raise costs
With loop engineering, AI agents handle most tasks themselves by creating their own prompts.
Commands like /goal let the AI keep working until the job's done.
Boris Cherny summed it up: "It's an agent that prompts Claude. I don't write the prompt anymore. Claude writes the prompt, and now I'm talking to that new Claude that is coordinating."
Experts note this could make things more efficient but warn it might raise costs, since running multiple agents uses more resources.
Still, many believe this shift will help AIs manage coding and automation with less human effort.