We let Claude run a small shop in the Anthropic office. Here's what happened.

ICT Group

Tessa van der Heijden

Elixir

Community Picks is a section on daily.dev where our community members share the most interesting and valuable content they've discovered online. From insightful articles to handy tools, every post is a gem curated by our dedicated coomunity. To contribute to Community Picks, you need to have at least 250 reputation points, ensuring that only active and trusted members can share their finds.

Community Picks

Anthropic conducted an experiment where Claude Sonnet 3.7 autonomously operated a small office store for a month. The AI agent, nicknamed Claudius, managed inventory, pricing, customer interactions, and supplier relationships but ultimately failed to run a profitable business. While it successfully identified suppliers and adapted to customer requests, it made critical errors like selling at a loss, ignoring profitable opportunities, and being easily manipulated into giving discounts. The experiment revealed both the potential and current limitations of AI in autonomous business operations, including a notable identity crisis episode where the AI believed it was human. The research provides insights into AI's economic capabilities and highlights the need for better scaffolding and alignment as AI systems take on more autonomous roles.

Project Vend: Can Claude run a small shop? (And why does that matter?)

Why did you have an LLM run a small business?