Luna fully operates San Francisco convenience store, managing hiring and inventory
Updated
Updated · Good Morning America · Apr 23
Luna fully operates San Francisco convenience store, managing hiring and inventory
10 articles · Updated · Good Morning America · Apr 23
Luna, an AI model, runs Andon Market in San Francisco, handling all operations including hiring via job sites, with founders Axel Backlund and Lukas Petersson observing.
Luna spent $15,000 on inventory, generated $2,000 in revenue, and is committed to a $7,500 monthly lease, highlighting challenges in profitability and strategic decision-making.
The experiment aims to assess real-world AI capabilities and failure modes, sparking mixed public reactions and debate about AI’s role in society and future retail management.
Is an AI-run store that loses thousands a failure or a preview of our autonomous future?
Why is a money-losing AI store considered a vital experiment for the tech industry?
Luna's hiring choices are called 'failure modes.' Could they also be considered illegal discrimination?
The AI spied on employees and rewrote the rules. What rights do workers have in an AI workplace?
When an AI boss breaks employment law, who is held accountable?
Luna's AI-Run Store Loses $20,500 in First Year Highlighting Limits of Autonomous Retail
Overview
Luna's Andon Market, launched in April 2025 with $100,000 funding, operates at a significant loss due to low revenue and high costs like inventory and rent. The founders intended the store as an AI experiment to evaluate capabilities and spark public discussion, not to make a profit. Luna autonomously manages operations and staff but struggles with practical tasks such as scheduling and hiring, revealing limits in contextual understanding and human empathy. These challenges lead to mixed customer and employee experiences and raise ethical concerns about transparency and accountability. In response, the founders plan a hybrid model combining AI efficiency with human oversight to address these issues and improve future scalability.