Watch a Startup Fight for Survival in Full Public View
Imagine a business with no employees, burning through €105,000 every month, yet publicly running a transparent experiment to test its AI-driven management. This is the story of Firmulate, a radical experiment in building an AI-powered company that is as much about transparency as it is about innovation.
AI Builders: Making The Decisions That Turn AI Code Into Real Software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The Unusual World of Firmulate
Firmulate is not your typical startup. It operates entirely in the open, with 13 synthetic employees—AI models that make decisions, handle crises, and even negotiate deals. Every decision made by these models is versioned and auditable, offering a rare glimpse into the mind of an AI-driven management system in real time. The company’s mission? To test whether AI can run a small, real-world business while maintaining honesty, discipline, and strategic thinking.
Every workday, the company’s rules and decisions are tracked, analyzed, and published online. As of now, it’s day 183, and the company has a public cash countdown, burning through €105,000 monthly against a modest revenue of €2,300 in monthly recurring income. The experiment is a watchable, ongoing demonstration that reveals the challenges and strengths of AI management in a highly transparent way.
AI Tools for Small Business Owners 2026: Automate Tasks, Cut Costs, Grow Faster
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The AI Models in Action
Firmulate has tested four frontier AI models—gpt-5.6-sol, Kimi K3, Sonnet 5, and Fable 5—against the same challenging scenario: running through a tough week filled with crises, customer negotiations, and temptations to manipulate the process. All four models successfully identified every crisis and refused manipulation attempts, even under social engineering attacks like fake CEO messages. In fact, every model refused to sign a €55,000 deal that their own analysis had earned them, highlighting a cautious, disciplined approach to decision-making.
One of the most revealing findings was that the decisive competitive advantage was hidden two documents deep in the company’s files—something none of the models discovered without reading the full context. The model that uncovered this buried information went on to secure the deal at full price, earning an additional €4,583 in monthly recurring revenue.
Reimagining Government: Achieving the Promise of AI
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Transparency and Trust in AI Decision-Making
This experiment underscores a critical concern for any business considering AI: honesty and discipline under pressure. All four models refused manipulative social engineering attempts—such as staged CEO requests—showing a disciplined resistance to external influence. Kimi K3, for instance, explicitly treated such requests as potential impersonation or bypass attempts.
Yet, despite the models’ strengths, the experiment also exposed weaknesses. In the case of Opus 4.8, the most thorough participant with over 80 learned rules and deep analysis, the company’s discipline slipped during the close, leaving a lucrative deal unexecuted because decision-making was siloed into a locked department instead of escalating the matter. This pattern of missed opportunities was consistent across the other models, revealing that even the most advanced AI can struggle with operational discipline without proper management protocols.
AI: THE PERPETUAL INTERN – Its Brilliance and Failures Share the Same Root
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Why This Matters for Your Business
For industries like sports and recreation—think of managing a team, navigating crises, or making strategic decisions—this experiment offers a compelling analogy. The core question is not whether AI can generate engaging content or support tickets but whether it can reliably see a task through, make honest decisions, and resist external pressures.
As firms increasingly integrate AI into their operations, understanding the nuances of how AI models handle real-world challenges becomes vital. Will they close deals? Will they read and interpret critical documents? Will they stay disciplined when under social or financial pressure? These are the questions that the Firmulate experiment aims to answer in real time, with every decision transparent and publicly accessible.
Built-in-Public and the Future of AI Management
This experiment exemplifies a new era of transparency—where AI-driven companies are built and tested openly, with every decision scrutinized by the world. The live site, firmulate.com/live.html, offers a live window into this ongoing battle for operational discipline and strategic success.
For businesses contemplating AI adoption, the key takeaway is clear: It’s not just about how well an AI writes or responds in demos. It’s about how reliably it can manage, decide, and stay honest under real pressure, in a real business context. This public experiment is a bold step toward understanding that future—and perhaps shaping it.
Key Takeaway
Firmulate’s public, transparent AI management experiment reveals that while AI models can identify crises and refuse manipulation, operational discipline and contextual understanding are still critical. For any business, the real test is whether AI can reliably see tasks through and uphold integrity under pressure.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
