What if your next AI assistant was running a real company — and losing €105,000 a month?
Imagine watching an actual business operate in real time, its decisions, struggles, and failures laid bare for all to see. That’s the premise behind a live experiment by Firmulate, where cutting-edge AI models manage a small software firm day by day, facing crises, making decisions, and even losing money — all publicly accessible at firmulate.com/live.
As an affiliate, we earn on qualifying purchases.
The Live Business in Action
Firmulate’s experiment is extraordinary in both scope and transparency. The company is run by 13 synthetic employees, each an AI model designed to emulate real management decisions. Every workday, these models handle the same set of crises — from customer disputes to internal miscommunications — with decision logs that are fully versioned and auditable. This setup allows observers to see precisely how each AI responds under pressure, whether it stays honest, and how effective its choices are.
Currently, the firm is losing €105,000 every month against a modest €2,300 monthly recurring revenue (MRR). Despite this, the live platform provides a rare window into the mechanics of AI decision-making in a complex, real-world scenario. Visitors can watch the company’s daily routines, read its self-learned playbook rules (more than 680), and even explore the decisions made on each day.
The Experiment’s Results
Four leading AI models, representing the frontier of language and decision-making systems, were tasked with managing the same tough week. All identified every crisis and refused every manipulation attempt, including social engineering tactics such as fake CEO messages and reporter tricks. However, only two of them managed to close the critical deal worth +€4,583 in monthly recurring revenue. The others, despite correct diagnosis, failed to follow through or left opportunities on the table.
Interestingly, the decisive factor was a buried piece of information in the company’s internal files — a detail that only the models that read deeper into the documentation managed to leverage. This allowed them to close the deal at full price, highlighting how important thorough data reading and understanding are in AI decision-making.
Why Does This Matter to Business and Tech?
For anyone managing a team, a support center, or a CRM, the key takeaway isn’t just whether an AI can generate convincing chat responses. It’s whether it can complete a task, stay honest under pressure, and dig into your internal data to find hidden opportunities. The experiment underscores a crucial truth: AI’s value in business isn’t just in language fluency but in integrity, diligence, and strategic depth.
As an affiliate, we earn on qualifying purchases.
The Company’s Unique Setup
The live operation is a true build-in-public experiment, with daily versions of its decision-making process publicly accessible. It’s a real company, with real money mechanics, risking €105k each month, and it’s fully transparent about its setbacks and successes. The platform also includes a quiz where 242 real, unedited management decisions are presented, challenging users to guess which AI model made each choice.
This openness aims to demonstrate the capabilities and limitations of current AI systems in operational contexts — a crucial perspective as AI tools become more integrated into business workflows.
As an affiliate, we earn on qualifying purchases.
What This Live Experiment Tells Business Leaders
Watching a real AI-managed company in action reveals the importance of thorough data reading, honest decision-making, and the ability to handle crises under pressure. The experiment shows that success isn’t just about spotting crises but also about following through and leveraging hidden information. For AI to truly add value to your business, it needs to do more than chat well — it needs to act reliably, ethically, and strategically, especially when stakes are high.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
As an affiliate, we earn on qualifying purchases.
