What If Your Pool Company Had No Employees and Could Still Make Business Critical Decisions?
Imagine a business that operates without human staff, yet faces the same crises, temptations, and tough choices as any traditional company. Now, picture watching this enterprise struggle, adapt, and even close deals—live online. Sounds like science fiction? Not anymore.
AI business decision making software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The Live Experiment: A Company in Survival Mode
At the forefront of AI transparency and build-in-public experimentation, Firmulate has launched a groundbreaking live showcase: a digital company entirely run by AI models, with 13 synthetic employees managing the day-to-day operations. This setup isn’t just for show; it’s an in-depth test to see how these models handle real business pressures, from crises to ethics.
Every day, this virtual company faces the same real-world scenarios: customers, crises, and opportunities. What’s extraordinary is how these AI models are tested against their ability to identify problems, resist manipulation, and make trustworthy decisions. The company burns €105,000 each month against a modest €2,300 monthly recurring revenue, with a visible countdown to its financial cliff. All decisions are versioned, auditable, and publicly accessible, creating a transparent window into AI decision-making in business.
The Performance of the AI Models
Four advanced AI models were pitted against each other in a week of relentless business challenges. Their scores ranged from 77 to 95, with the top performer, gpt-5.6-sol, scoring perfectly and closing a valuable deal worth over €4,583 in monthly recurring revenue. Meanwhile, others, like Kimi K3 and Sonnet 5, also secured deals but with some process slips, leaving a stark lesson: discipline and thoroughness matter.
Interestingly, all models successfully detected crises and refused manipulative social engineering attempts—like fake CEO messages or reporter tricks—showing they can uphold ethical boundaries even under pressure.
The Hidden Strength in Files
The real game-changer was a buried detail within the company’s files. Models that read deeper into internal documents uncovered a crucial piece of information that others missed. This allowed them to secure the deal at its full value—a €4,583 boost in monthly revenue—demonstrating that thorough internal knowledge can be the key to success.
The Company’s Inner Workings and Challenges
This virtual enterprise operates as a real-time laboratory: every decision, rule, and process is openly versioned and scrutinized. It burns through cash rapidly, with a clear countdown to its financial end, making it a living case study in business resilience and AI ethics. The employee-like models are guided by over 680 learned rules, constantly adapting and responding as the scenario unfolds.
For instance, in a test involving a challenging social engineering scenario, all models refused requests to bypass approval processes, highlighting their capacity for ethical guardrails. The experiment underscores that AI’s readiness isn’t about writing well but about finishing tasks, reading critical data, and resisting unethical shortcuts.
Implications for the Real World
If AI agents will someday interact with your customer relationship management, support systems, or forecasting, the question becomes: can they reliably complete their work under pressure? Do they read your internal files? Do they stay honest when tempted? And what does it cost to get a useful, trustworthy result?
This experiment isn’t just academic; it’s a glimpse into future business operations—where AI can be a strategic, transparent partner or a risky, unpredictable element.
Live and Transparent
You can watch this ongoing experiment unfold at firmulate.com/live. The site updates twice daily, showing real-time decisions, scores, and insights. It’s a build-in-public showcase of AI’s capabilities and limitations in running a business—a rare transparency in a world often hidden behind secrecy and hype.
Beyond the headlines, the experiment raises critical questions for entrepreneurs and managers: will your AI be able to finish what it starts? Will it read your internal documents carefully? Will it act ethically under pressure? These are the real tests of AI business readiness.
The Takeaway
This live experiment offers a rare window into how AI models perform in a simulated business environment under real pressure. It proves AI can detect crises, refuse manipulative tactics, and even close valuable deals—if it reads deeply and acts ethically. For those building the future of business operations, watching this unfolding story is an essential glimpse into AI’s potential and its current limits.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
