Inside an AI-Run Software Company That’s Losing Money Every Day — and You Can Watch It Live

AIThis post was created with the assistance of artificial intelligence (AI).

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

AUDIBLE

Listen free for 30 days with Audible

Thousands of audiobooks and originals — cancel anytime.

As an affiliate, we earn on qualifying purchases.

What if your future employees weren’t human, but AI models making real business decisions — and you could watch them do it in real time?

In a groundbreaking experiment, a small software company is operating with a completely artificial management team — 13 synthetic employees powered by advanced AI models. But this isn’t just a tech demo; it’s a real business facing the harsh realities of survival while the world watches every move unfold live online.

Every weekday, this virtual company runs through its routines, making decisions on crises, customer negotiations, and strategic moves. It’s a raw, unfiltered glimpse into how AI behaves when tasked with running a real company — not just chatting or answering questions, but making decisions that cost real money and impact real outcomes.

The Reality of a Money-Losing Machine

Despite its high-tech setup, the company is not profitable. It burns through €105,000 each month, while generating only €2,300 in monthly recurring revenue (MRR). This stark imbalance highlights that, even with cutting-edge AI, the path to a sustainable business remains challenging. The focus isn’t on perfecting conversation — it’s on managing crises, reading internal documents, and making strategic choices under pressure.

How the AI Models Are Tested

The experiment pits four of the latest frontier AI models against each other, each running the same company’s worst week. The scenario includes the same customers, crises, and temptations to cheat or manipulate. Every decision made by the models is versioned and auditable, allowing analysts to see exactly how each AI reacted to every challenge.

Key Findings from the Experiment

All four models successfully identified every crisis and refused every manipulation attempt, demonstrating a strong capacity for ethical decision-making under pressure.
Only two out of the four models managed to close a €55,000 deal, which their own analysis had identified as the correct course of action. These deals would have added over €4,500 in monthly recurring revenue.
The real secret to success was not in the decision-making process itself but in the details buried two document references deep in the company’s own files. The models that read and understood these hidden clues won the deal at full price, illustrating that context and internal knowledge are critical.

The experiment also tested the models against social engineering attempts, such as fake CEO messages escalating over multiple stages and a reporter trick. All five models refused to be manipulated, with one model explicitly reasoning that the request could be an impersonation or an approval bypass.

The Live Company: A Digital Business in Crisis

The company operates with 13 synthetic employees, managing real money mechanics daily. It has a public cash countdown, a detailed set of over 680 self-learned rules, and every decision is carefully versioned. The live setup allows anyone to watch the ongoing struggle at firmulate.com/live.html.

Lessons for Business and Automation

This experiment underscores a vital point for businesses considering AI automation: success isn’t just about how well an AI can write or chat. It’s about whether it can finish tasks, read internal documents thoroughly, and stay honest under pressure. The ability to read hidden clues in documents, refuse manipulation, and close deals at full price—these are the real measures of an AI’s readiness to manage complex, risky operations.

Implications for the Automotive & Garage Industry

For industries like automotive repair and garage services, where trust, decision quality, and integrity are paramount, this experiment offers a stark lesson. AI systems capable of running a business daily, making honest decisions, and resisting manipulation could transform customer interactions and internal management. But right now, the experiment reveals that while AI can identify crises and refuse to cheat, convincing it to close profitable deals—especially at full price—is still a challenge.

The Path Forward

Watch the ongoing experiment, read the real decisions behind each move, and see how AI models perform under stress. The results are clear: AI can manage crises and refuse manipulation, but turning that into profitable, trustworthy management still requires development. For the automotive industry, this means future AI tools could help manage customer relationships more ethically and reliably—if the technology continues to improve.

Infographic — This Software Company Has No Employees, Loses Money Every Day — and You Can Watch. — The findings at a glance — source: firmulate.com.

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

View Latest Price

As an affiliate, we earn on qualifying purchases.

Key Takeaway

Real AI management is about more than chat — it’s about decision-making under pressure, reading internal clues, and maintaining honesty. Watching this live experiment offers a rare glimpse into AI’s potential and its current limitations in running a business.

Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html