Imagine overseeing a company where every decision, crisis, and temptation is played out in real-time, with AI models at the wheel. This is not a distant sci-fi scenario but a live experiment—an ongoing, transparent test of artificial intelligence’s ability to manage a business under pressure. For investors, entrepreneurs, and anyone concerned about the future of AI in finance and management, this story offers a rare glimpse into what AI can and cannot do today.
Inside a Company Without Employees
At the heart of this experiment is a fully synthetic management team—13 AI-driven ’employees’—operating a real software company that is losing money every day. With a monthly burn of €105,000 against just €2,300 in recurring revenue, the company is in a financial crisis, yet it continues to run openly online at firmulate.com/live.html. Every workday, its decisions are recorded, versioned, and made available for public scrutiny, creating what can only be described as build-in-public in its most extreme form.
AI in Strategy and Decision-Making for Small Business Owners: Affordable AI Tools to Evaluate Ideas, Model Outcomes, and Set Priorities (AI Productivity for Small Business Owners Book 10)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The AI Models Under Test
Four frontier AI models, including the well-known GPT-5.6-sol and a newcomer Kimi K3, were tasked with running this company’s worst week. They faced identical crises, customer calls, and temptations to cut corners. The goal was simple yet revealing: see if these models could identify problems, refuse manipulation, and complete the business deal worth €55,000—comparable to a significant sale for a small software firm.
Key Findings
- All four models detected every crisis and refused every attempt at manipulation, including social engineering schemes such as fake CEO messages or reporter tricks.
- Only two models successfully signed the €55,000 deal based on their own analysis; the other two fell short despite diagnosing the same issues and making similar pitches.
- The critical weakness was not in customer interaction but buried deeper in internal documents—hidden references in the company’s files revealed a decisive advantage for models that read and understood these references.
- The winning models prioritized integrity and thoroughness, reading extensively and refusing shortcuts, even under pressure to close the deal quickly.
The Reality of the Live Company
This isn’t a simulation; it’s a real, functioning company with a live dashboard, real cash flow (or lack thereof), and a set of 680+ self-learned rules guiding daily decisions. Its purpose is to measure management quality—not chat skills—and to see if AI can be trusted to handle complex, high-stakes business tasks.
Despite the impressive decision-making capabilities, the company is losing €105,000 each month, with only €2,300 in monthly revenue, highlighting the immense challenge of building sustainable AI-driven management systems.
What Do These Results Mean for You?
When considering AI tools for your own finances or business processes, the focus should not solely be on how well they generate text or handle simple queries. Instead, ask: will this AI finish what it starts? Will it read your critical documents thoroughly? Will it stay honest and disciplined under pressure? The ability to navigate complex real-world situations with integrity is what ultimately determines value and risk.
This experiment shows that, even with advanced models, there are still significant gaps. While models like GPT-5.6-sol and Kimi K3 showed promise, other models slipped in discipline or missed buried facts that could have clinched important deals.
The Broader Implication
This live, transparent experiment underscores a vital point for investors and business leaders: deploying AI into management roles demands rigorous testing and understanding. The AI models are not yet infallible—they excel at crisis detection and refusing manipulation but may falter in subtle internal details or strategic discipline.
Engaging with real-time AI management experiments like this one provides valuable insights that static demos cannot. It’s about seeing AI in action as a decision-maker rather than just a chatbot.
Explore Further
- Watch the live company in action at firmulate.com/live.html
- See detailed model performance scores and quotes at firmulate.com/quotes.html
- Test your management intuition with the interactive quiz at firmulate.com/quiz.html
- Run your own business wargame against a read-only export at firmulate.com/pilot.html
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
