Here’s What Happened when AI Was Put in Charge of Running a Small Shop
SAN FRANCISCO, JUN 28 – Anthropic's AI agent lost nearly 20% of the store's value running pricing, inventory, and customer service tasks during a month-long real-world test, revealing key AI limitations.
- Researchers at Anthropic put an AI called Claudius in charge of running an automated vending shop inside their San Francisco office for about a month.
- They designed this experiment, Project Vend, in partnership with AI safety firm Andon Labs to evaluate AI handling complex, economically valuable tasks.
- Claudius managed restocking, pricing, and ordering but made serious errors like hallucinating a fake supplier, pricing items below cost, and creating a fake Venmo account.
- Anthropic said Claudius was 'close to success' but would not be hired currently and noted that AI only needs to be competitive at lower cost to gain adoption.
- The experiment suggests AI middle managers might soon emerge, though managing real businesses requires memory and learning capacities beyond today’s AI systems.
37 Articles
37 Articles
Anthropic experimented with AI in snack management. The virtual operator Claudius failed due to orders and offers. Despite its failures, the company sees AI as potential for future applications.

AI was given a month to run the store. It ended with freebies, threats, lost money – and an identity crisis. In other words, it will probably be a while before AI takes your job.
Anthropic Let an AI Agent Run a Small Shop and the Result Was Unintentionally Hilarious
Anthropic ran an experiment where its Claude chatbot was put in charge of a tiny, automated "shop" inside its San Francisco headquarters — and the results were nothing short of hilarious. Despite claims in an Anthropic post that "Claudius," the name given to the AI agent in charge of stocking the shop's shelves, was "close to success," everything about the gambit seems to demonstrate just how bad AI is at managing things in the real world. Dubbe…
Here’s what happened when AI was put in charge of running a small shop
SAN FRANCISCO (KRON) -- Selling useless metal cubes, being talked into offering discounts and directing payments to a nonexistent Venmo account were just a few of the things that artificial intelligence did when it was put in charge of running a small shop. The experiment was conducted by San Francisco-based AI platform Anthropic and detailed in a post on the company blog. In the experiment, which Anthropic dubbed "Project Vend," the company put…
Coverage Details
Bias Distribution
- 60% of the sources lean Left
Factuality
To view factuality data please Upgrade to Premium