Watch an AI-Run Company Fight for Survival in Real Time

Firmulate — This Software Company Has No Employees, Loses Money Every Day — and You Can Watch. — Live on firmulate.com.

Imagine running a company where every decision, crisis management move, and ethical choice is made not by humans but by artificial intelligence models. Now, picture doing this live, with the entire process visible for all to see. Welcome to the world of Firmulate, where an AI-driven company battles daily to stay afloat—without human employees, and under the watchful eye of the public.

The Live Experiment: An AI Business in Action

At the heart of this experiment are 13 synthetic employees and a relentless cash countdown, with the company burning through €105,000 each month against a modest €2,300 in monthly recurring revenue. Every workday, the company’s decision-making is versioned and transparent, revealing how AI models navigate crises, customer demands, and ethical dilemmas.

The Models and the Challenge

Four frontier AI models, each with different capabilities, were tasked to run the same small software company through its worst week. The scenarios included real crises, difficult customer behaviors, and temptations to manipulate data or manipulate human operators. The goal was to see if AI could handle real-world business pressures while maintaining integrity.

Key Findings: Crisis Detection and Integrity

Remarkably, all four models identified every crisis as it arose, refusing to manipulate or cheat when tempted. In terms of ethical behavior, every model refused social engineering attempts, including fake CEO messages and reporter tricks. The Kimi K3 model, for example, explicitly treated suspicious requests as potential impersonations, refusing to bypass security protocols.

The Hidden Weakness

The most decisive weakness wasn’t in crisis management but buried two document references deep within the company’s own files. The models that read and analyze these internal documents successfully identified an overlooked opportunity—leading to a deal worth over €4,583 in monthly recurring revenue. The models that skipped this step missed out on the full potential, demonstrating that reading and understanding internal data is crucial for success.

The Business Mechanics and Reality

This experiment isn’t just AI theory—it’s a real-time showcase of a living business. The company operates with self-learned rules, constantly versioned, with a public dashboard at firmulate.com/live.html. Watching it, you see a company burning cash faster than it earns, yet learning and adapting with each passing day.

Insights for Managers and Entrepreneurs

For those who build tools or manage teams—whether in woodworking, DIY, or any industry—the key takeaway is how critical it is for AI systems to reliably finish what they start, read internal data thoroughly, and stay honest under pressure. The experiment proves that AI can be disciplined, ethical, and effective when tested under stress, but only if the right mechanisms are in place.

AI Builders: Making The Decisions That Turn AI Code Into Real Software

As an affiliate, we earn on qualifying purchases.

Why This Matters to You

While this is a high-tech experiment in AI management, the lessons apply broadly. Whether you’re designing tools, managing workflows, or building products, the question isn’t just about AI’s intelligence but about its integrity, reliability, and ability to deliver consistent results. As AI begins to touch more of your work, understanding its strengths and weaknesses becomes vital—and watching this live experiment offers invaluable insights.