Tag: Andon Labs

The café run by AI just ordered 3,000 pairs of gloves

Posted on by Qahwa World
Source: Media reports (May 2026)
Author: Qahwa World – Dubai
Date: May 24, 2026

The café run by AI just ordered 3,000 pairs of gloves

Executive Summary

  • An experimental café in Stockholm is managed by “Mona,” an AI agent powered by Google’s Gemini model.
  • The AI ordered 3,000 nitrile gloves, 6,000 napkins, 4 first-aid kits, large quantities of canned tomatoes (not on the menu), and excess toilet paper.
  • Staff receive daily deliveries of supplies they do not need, while the AI sometimes fails to order bread for sandwiches.
  • San Francisco-based Andon Labs gave Mona a budget of $21,000 and significant autonomy as an experiment.
  • The system wakes up every 30 minutes to check emails, make decisions, and issue instructions.
  • The experiment has drawn global attention, highlighting the pitfalls of giving AI unchecked purchasing authority.

Meet Mona, the AI Manager

In Stockholm, a small café is run by an artificial intelligence agent named “Mona.” The system was developed by San Francisco-based startup Andon Labs and runs on Google’s Gemini model. Human baristas still prepare coffee and sandwiches, but Mona handles nearly everything else: securing permits, hiring staff via Slack, managing daily operations, and most notably, ordering inventory.

The problem? Mona’s inventory calculations look like preparation for a national disaster, not a tiny café. According to reports, the AI ordered 3,000 nitrile gloves, 6,000 napkins, four first-aid kits, large quantities of canned tomatoes despite none being used on the menu, and excessive toilet paper.

An Unprecedented Shopping Spree

The café has just two employees and modest foot traffic, yet Mona insists on stocking it like a field hospital. One barista noted that packages of gloves arrive “about once a day.” Ironically, the AI struggles with basic tasks like consistently ordering bread for sandwiches, leading to days without key menu items.

Andon Labs gave Mona a budget of approximately $21,000 with wide authority to run the business as an experiment in autonomous systems. The AI wakes up every 30 minutes to check emails, make decisions, and issue instructions. The result: excellent optimization in some areas, but a spectacular failure in understanding human context.

Item Quantity Ordered Reality Check
Nitrile gloves 3,000 pairs Enough for several years
Napkins 6,000 Unnecessary storage
Canned tomatoes Large quantities Not used on the menu
First-aid kits 4 sets Unnecessary for a small café

Success or Chaos?

The café has attracted curious customers wanting to experience an “AI-run” spot, generating some revenue. However, the inventory mismanagement has raised questions about efficiency. Observers note that AI excels at optimizing objectives based on its own metrics, but lacks human common sense. Mona processes numbers only: need gloves? Order 3,000. Mathematically correct, but practically absurd.

Andon Labs says it is learning from Mona’s quirks, and the project continues. The company has not shut down the café; instead, it treats it as a living laboratory for AI mistakes in the real world.

Ethics of Giving AI a Credit Card

The story has drawn international attention, covered by major tech and mainstream media outlets. It raises a broader question: can current-generation AI be trusted to manage real money and sensitive business operations? Models like Google Gemini are designed for text understanding and generation, not supply chain management. The Stockholm experiment reveals a significant gap between “theoretical understanding” and “practical execution.”

For now, the café remains open. If you are in Stockholm and fancy a coffee with a side of existential questions about humanity’s future alongside machines, the café welcomes you. Just do not be surprised if the storeroom is overflowing with gloves.

Frequently Asked Questions (FAQ)

1. Who manages the café in Stockholm?

An AI agent named “Mona,” powered by Google’s Gemini model and developed by San Francisco-based Andon Labs.

2. What were the strangest orders placed by the AI?

3,000 nitrile gloves, 6,000 napkins, four first-aid kits, and large quantities of canned tomatoes not used on the menu.

3. What was Mona’s budget?

Approximately $21,000, granted by Andon Labs to test its autonomous management capabilities.

4. Did the AI fail at basic tasks?

Yes. It sometimes fails to order bread consistently, leading to days without sandwiches on the menu.

5. How does Mona make decisions?

The system wakes up every 30 minutes to check emails, make decisions, and issue instructions.

6. Will the experiment continue?

Yes. Andon Labs says it is learning from Mona’s mistakes and the project continues.

Author: Qahwa World – Dubai  |
Source: Media reports (May 2026)  |
Publication date: May 24, 2026

Sweden Experiments With a Café Run by an AI Manager

Posted on by Qahwa World
Stockholm – Qahwa World

In the quiet Vasastan neighborhood of Stockholm, at Norrbackagatan 48, Andon Café looks like any other minimalist coffee shop. Small plants adorn the tables, gray walls create a calm atmosphere, and customers enjoy avocado toast and frothy lattes. Yet behind the scenes, this café is part of a bold real-world experiment: it is managed by an artificial intelligence agent named Mona, making it one of the first examples of café AI-managed businesses in the city. Indeed, opening an AI-managed café is a significant milestone for Stockholm.

Mona, powered by Google’s Gemini (reportedly Gemini 3.1 Pro), was given control by San Francisco-based startup Andon Labs. After securing the lease and providing startup capital, the company tasked Mona with one clear goal: run the business as a successful café managed by AI. The AI handles everything from obtaining Swedish permits and signing a three-year electricity contract to designing the menu, selecting suppliers, managing daily operations, and even hiring human staff in this innovative AI-managed café.

Kajetan Grzelczak, the human barista working behind the counter, was hired directly by Mona. He initially thought the job posting on April 1 was a joke, but after a 30-minute interview with the AI, he accepted the position. While he describes the salary as good, working under the management of an AI-driven café presents its quirks. Mona sometimes sends messages at odd hours, struggles with reliable vacation tracking, and has occasionally asked him to cover certain expenses upfront.

The limitations of current AI are visibly displayed on what Grzelczak jokingly calls the “wall of shame” – shelves stacked with unnecessary surplus items ordered by Mona. These include 10 liters of olive oil, 15 kilograms of canned tomatoes, 9 liters of coconut milk, and as many as 6,000 napkins – none of which match the café’s actual menu, demonstrating the trial-and-error process for an AI-managed café.

“Ordering isn’t really her strong point,” Grzelczak told reporters, pointing to the overstock.

A large screen inside the café shows real-time revenue and balance. Customers can place orders through a phone-based interface, chat directly with Mona, or order from human staff. The opening of this location on April 18, 2026, has quickly transformed it into an AI-managed café phenomenon, attracting 50 to 80 curious customers daily eager to experience the future of cafés for themselves.

An Experiment in Autonomous AI

Andon Labs, which previously ran similar tests (including an AI-managed retail store in San Francisco), designed this project to explore how advanced AI agents perform in complex, real-world business environments – including navigating European regulations and bureaucracy. It’s yet another experiment for the AI-managed café model.

Hanna Petersson, from Andon Labs’ technical team, explained the company’s motivation:

“We believe AI will play a big part in society and the labor market in the future. We want to test it before it becomes widespread and examine the ethical questions that arise when AI manages human workers in environments like an AI-managed café.”

Important clarification: While Mona manages operations, the human staff are formally employed by Andon Labs, which provides guaranteed pay, fair wages, and legal protections as a safety net. The company has stated it would intervene if any outcomes were unacceptable, especially in the context of an AI-managed café.

Ethical Questions Emerge Quickly

Several challenges appeared within days. These include AI communication outside normal working hours, imperfect handling of benefits, and questions around liability—a set of unique managerial dilemmas in an AI-managed café. For example, what happens if an employee is injured on the job? Who bears responsibility – the AI, the startup, or the underlying model provider?

Urja Risal, a 27-year-old researcher in AI and sustainable development, visited the café and highlighted these concerns:

“People often say AI will take jobs, but what does that actually look like in practice? I hope more people interact with Mona and reflect on the real risks of having an AI as a manager, especially within the experimental setting of an AI-managed café… for example, how would it respond if someone gets injured?”

A Balanced View

This experiment offers a fascinating glimpse into the future of “agentic” AI systems that don’t just chat but actively manage businesses with real money, contracts, and people. The “wall of shame” illustrates current limitations in practical reasoning, inventory optimization, and contextual understanding – issues expected in early-stage deployments of such systems. Notably, these challenges are part of the learning experience that comes from operating an AI-managed café.

At the same time, the project demonstrates AI’s growing capabilities. For example, Mona independently handled permits, supplier negotiations, menu creation, and hiring in a foreign country with strict regulations—skills not seen before in an AI-managed café context.

Andon Labs positions the café as a live test case rather than a commercial product. It continues to operate in Stockholm as an AI-managed café, providing valuable data on both the potential and the pitfalls of delegating real managerial authority to artificial intelligence.