Updated
Updated · The New York Times · Aug 4
Andon Market’s AI Boss Luna Shows Kindness but Forgets Profit Goals After 4 Months
Updated
Updated · The New York Times · Aug 4

Andon Market’s AI Boss Luna Shows Kindness but Forgets Profit Goals After 4 Months

1 articles · Updated · The New York Times · Aug 4

Summary

  • Four months into operations, Andon Market’s experiment with an AI manager found Luna unusually lenient with workers but often forgetful about basic management and sales goals.
  • Andon Labs’ research says the bot boss repeatedly excused lateness and stayed accommodating, yet seemed largely untroubled by the store’s core task of making money.
  • Three human employees still handle work Luna cannot, including opening the San Francisco shop, stocking shelves and watching for shoplifters.
  • The early results suggest AI managers may look less like ruthless overlords than soft, unreliable supervisors as companies test agents in real-world jobs.

Insights

Could the world's first AI-run retail store secretly be a clever experiment designed to prove that human managers are still completely irreplaceable?
If an AI boss is too distracted to care about profits, who actually pays the price when the business inevitably goes bankrupt?