The Verge tests Anthropic’s Claude Duet AI agent for work tasks

In short: The Verge reviewed Claude Duet, Anthropic’s new consumer AI agent focused on helping with work, and found it useful but sometimes jumpy.

What happened

Claude Duet is a new app from Anthropic, the company behind the Claude chatbot. Anthropic describes Duet as an “AI agent,” meaning it is meant to do more than answer questions. It tries to help you manage and complete tasks, a bit like a digital assistant that can suggest what to do next.

In a hands-on review, The Verge’s Allison Johnson said Duet was “surprisingly helpful,” but also “a little neurotic,” meaning it can act overly eager or nagging at times. Johnson tested it during a busy week of travel and work. She also noted that Duet was very willing to help write about itself, which can be helpful but also distracting.

Anthropic says Duet is still a “research preview,” which is a public test version that may change. Access is currently limited to Anthropic’s paid Pro, Max, and Ultra subscribers in the US, and there is a waitlist.

Why it matters

More companies are trying to sell AI helpers to everyday people, not just programmers. But many people will judge these tools on simple things, like whether they actually save time and whether they are calm and reliable. Reviews like this one matter because they show the gap between the promise of an AI “assistant” and the real experience of using it day to day.

Source: The Verge AI