How shifting risk to users makes Claude Code more powerful

People are discovering that Claude Code isn’t just for code.

Anthropic’s Claude Code has been gaining popularity among programmers since its launch last February. When I first wrote about the tool back in May, it was little known among non-programmers.

That started to change over the holidays. Word began to spread that — despite its name — Claude Code wasn’t just for code. It’s a general-purpose agent that can help users with a wide range of tasks.

Claude Code is “marketed as a tool for computer programmers, so I wasn’t using it because I’m not a computer programmer,” wrote the liberal Substack author Matt Yglesias on December 26. “But some friends urged me to fire up the command line and use it.”

“In a sense, everything you can do on a computer is a question of writing code,” Yglesias added. “So I downloaded the entire General Social Survey file, and put it in a directory with a Claude Code project. Then if I ask Claude a question about the GSS data, Claude writes up the R scripts it needs to interrogate the data set and answer the question.”

Last week, Anthropic itself capitalized on this trend with the release of Anthropic Cowork, a variant of Claude Code designed for use by non-programmers.

Claude Code is a text-based tool that runs in a command-line environment (for example, the Terminal app on a Mac). The command line is a familiar environment for programmers, but many normal users find it confusing and even intimidating.

Cowork is a Mac app that superficially looks like a normal chatbot. Indeed, it looks so much like a normal chatbot that you might be wondering why it’s a separate product at all. If Anthropic wanted to bring Claude Code’s powerful capabilities to a general audience, why not just add those features to the regular Claude chatbot?

What ultimately differentiates Claude Code from conventional web-based chatbots isn’t any specific feature or capability. It’s a different philosophy about risk and responsibility.