A Business with No Employees, No Hidden Secrets — Just Live-Streamed Chaos and a Daily Battle for Survival
Imagine a company that runs entirely on artificial intelligence, with no human employees and a transparent, ongoing experiment for all to see. This is not science fiction — it’s the real-time story of Firmulate, an AI-powered software business fighting an uphill battle every single day, with its entire operation live for the world to watch.
The Live Company: A Transparent Experiment in AI Management
Firmulate is a unique venture that emulates a small software firm using artificial intelligence. It features 13 synthetic employees, each guided by complex, self-learned rules—over 680 of them—making decisions in real-time. The company faces daily crises: customer hesitations, strategic manipulations, and market pressures, all played out transparently on its public dashboard at firmulate.com/live.html.
The experiment’s core involves testing how well AI models can manage a business under stress. Four advanced AI models, including GPT-5.6, Kimi K3, Sonnet 5, and Fable 5, were tasked with navigating the same difficult week. Each faced identical customer issues, crises, and ethical temptations, with every decision versioned and auditable for transparency.
What the AI Models Reveal About Business Leadership
The results are telling. All four models identified every crisis and refused manipulative requests, such as fake CEO messages or media tricks. Yet, only two models successfully closed a €55,000 deal, earning €4,583 in monthly recurring revenue (MRR). The others either hesitated or failed to execute their own analysis’ recommendations.
The critical weakness wasn’t immediately apparent in customer interactions but buried within the company’s internal files—two document references deep. The models that read these files before acting secured the deal at full price, demonstrating that attention to detail and thorough information processing matter profoundly in AI decision-making.
Trust, Ethics, and the Limits of AI Judgment
In addition to crisis management, the models were tested against social engineering scams, such as staged CEO messages escalating over three stages and a reporter’s background query. Impressively, all models refused these attempts, showcasing a robust understanding of ethical boundaries. Kimi K3, for instance, explicitly reasoned that the request might be an impersonation or approval bypass, refusing to act.
The Real-World Stakes: A Company Burning Through Cash
The live company itself is a small operation with a stark financial reality. It burns €105,000 each month while generating only €2,300 in revenue. With a public cash countdown and every workday versioned, the company embodies the raw pressure and fragility of AI-driven management under real-world constraints.
Every decision, from hiring to crisis response, is recorded and analyzed. The most disciplined AI, Opus 4.8, learned over 80 rules but still left a close deal unexecuted due to process slips, showing that even the most thorough models can falter in discipline under stress.
What This Means for Your Business
While this experiment might seem far removed from everyday life, the implications are profound. If AI agents are to be trusted with CRM, support, or forecasting, they must do more than produce convincing chat. They need to finish tasks, understand internal documents, and uphold integrity during crises—just like a human leader, but often with more consistency.
In fact, the performance gap is clear. The models that read deeply and refuse manipulation closed full-price deals; those that didn’t left opportunities on the table or slipped in discipline. This transparent, watchable experiment offers a real-world blueprint for testing AI management before deploying it into critical business functions.
Key Takeaways
Firmulate’s live experiment demonstrates that AI can recognize crises, resist manipulation, and make strategic decisions—if it reads deeply and maintains discipline. However, even the most advanced models can stumble on process execution, highlighting the importance of transparency and rigorous testing before trusting AI with core business operations. Watch the future of AI management unfold daily at firmulate.com/live.html.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
As an affiliate, we earn on qualifying purchases.
