Imagine running a small business where every decision is made by artificial intelligence, facing real crises, temptations, and financial struggles — all in full public view. This is not a science fiction scenario but the bold experiment of Firmulate, a live company powered entirely by AI models, where every move is observable and every mistake visible.
The Unique Live Experiment: An AI-Run Company in Action
Firmulate presents a groundbreaking experiment: a real, functioning software company run daily by four different frontier AI models. Unlike typical demos or chatbots, this company is a fully operational entity, complete with 13 synthetic employees, real money mechanics, and a public countdown to its cash reserves. The company burns through €105,000 each month while earning €2,300 in monthly recurring revenue, making its financial struggles starkly visible.
Every day, the AI models face the same set of crises — customer issues, internal dilemmas, and ethical tests — with decisions recorded and made transparent for all to see. This setup allows observers to evaluate not just what the AI can say, but what it can do: can it recognize critical issues, avoid manipulation, and close deals honestly?
The Results: AI Models Stand Up to Real-World Pressures
The experiment’s results are revealing. All four AI models successfully identified every crisis and refused every manipulation attempt — a crucial test of honesty and integrity. For instance, the models faced a staged social engineering attack involving fake CEO messages and a reporter’s subtle pressure. Remarkably, all five models refused to engage with the manipulation, demonstrating a strong ethical stance driven by their training.
However, when it came to closing a deal, only two models, GPT-5.6-sol and Kimi K3, managed to secure the agreement worth €55,000—a significant milestone for this AI-driven venture. The other two, including the most thorough participant Opus 4.8, identified the opportunity but left the deal on the table, illustrating how process discipline and decision-making nuances matter in such high-stakes environments.
As an affiliate, we earn on qualifying purchases.
The Hidden Weaknesses and What They Reveal
While the models demonstrated technical competence, the real weakness lay in their reliance on internal documentation. A crucial competitive advantage lay buried in the company’s own files—information that the models which read deeper into the files discovered, leading to successful deal closure at full price. This highlights an important insight: AI’s effectiveness heavily depends on access to the right information.
Build-in-Public in Its Most Extreme Form
This live experiment is not just a showcase; it’s a transparent window into the ongoing struggle of a small business under AI management. It is openly accessible at firmulate.com/live.html, where viewers can watch the company’s daily operations, read real decisions, and observe the financial drain firsthand.
enterprise AI decision-making tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Implications for the Future of AI in Business
What does this experiment tell us about the future of AI in enterprise management? For one, it underscores the importance of trustworthiness and thoroughness. The models’ ability to recognize crises and resist manipulation is promising, but their failure to fully close deals or escalate issues shows there’s still a significant gap to close before AI can fully replace or augment human decision-makers.
Moreover, the experiment emphasizes that success isn’t only about writing convincingly in chat; it’s about completing meaningful work—reading critical information, staying honest under pressure, and executing complex processes reliably. As AI models become more integrated into business operations, these qualities will be crucial for their effectiveness and integrity.
As an affiliate, we earn on qualifying purchases.
How You Can Engage
Interested in testing AI’s potential for your own enterprise? Firms can run their own version of the experiment through a read-only ‘wargame’ that simulates their business environment without risking real systems. This allows organizations to evaluate how AI models handle crises, ethics, and decision-making in a controlled, transparent setting. More details are available at firmulate.com/pilot.html.
Why This Matters for Everyone
Whether you’re a kitchen appliance enthusiast or a business leader, the core lesson is the same: AI’s true value lies in its ability to reliably execute tasks and uphold integrity, especially under pressure. The Firmulate experiment puts this to the test in real time, showing that AI can identify problems and stand firm against manipulation — but closing deals and making strategic decisions still require a nuanced approach.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
As an affiliate, we earn on qualifying purchases.
