Get baking supplies delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
What if your favorite bakery ran entirely on AI—no employees, no human hands, just algorithms?
Imagine a bakery that operates 24/7, makes decisions on ingredient sourcing, customer interactions, and crisis management—all mediated by artificial intelligence. You might wonder: can AI truly handle the complex, unpredictable world of running a business? The answer is being tested in real-time, with a groundbreaking, transparent experiment that you can watch live.
As an affiliate, we earn on qualifying purchases.
The Live Company That Runs on Algorithms
At the forefront of this exploration is Firmulate, a company that has set up a remarkable project: a real, functioning small software business operated entirely by AI models. Unlike typical demos or simulated test cases, this is a fully live operation, with every decision, crisis, and outcome publicly accessible at firmulate.com/live.html.
This company employs 13 synthetic ’employees’—AI models designed to mimic management decision-making—who face the same challenges as human managers: market crises, customer demands, internal errors, and temptations to cut corners. The twist? Every workday is versioned, every decision is auditable, and the entire process is open for scrutiny.
AI decision-making tools for small business
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The Experiment: Testing Different AI Models in the Same Crisis
The core of this experiment involves four next-generation AI models—each given the same small business and subjected to the same tough week filled with real crises and ethical tests. These models include GPT-5.6, Kimi K3, Sonnet 5, and Opus 4.8, each with distinct capabilities and approaches.
All models successfully detected every crisis they encountered and refused every manipulation attempt, including social engineering tactics like fake CEO messages and reporter tricks. This indicates a high level of integrity and trustworthiness, crucial qualities for automation in sensitive business operations.
AI deal-closing automation software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Can AI Close Deals? The Critical Difference
Despite their shared competence in crisis detection and resistance to manipulation, only two models managed to close a key deal worth €55,000 monthly recurring revenue—that is, they signed the contract resulting from their own analysis and recommendations. The other two, despite identical diagnoses and pitches, walked away without sealing the deal.
This discrepancy highlights an important fact: the decisive leverage was hidden within the company’s own documentation. The models that read deeper into internal files uncovered critical information, enabling them to win the deal at full price—an advantage of over €4,500 MRR. It shows that even in AI decision-making, thorough information processing makes the difference between success and failure.
As an affiliate, we earn on qualifying purchases.
Building Public, Building in Public
This experiment is as much about transparency as it is about AI performance. Every decision made by each model is versioned and published, making it possible to see how they analyze, prioritize, and respond to crises. The platform also hosts a quiz where viewers can guess which model made which decision, engaging the audience in this real-time management battle.
Real Money, Real Stakes
Beyond the AI models, the ‘live company’ itself is a real business—burning €105,000 every month against a modest €2,300 in monthly revenue. This stark financial reality underscores the urgent need for effective automation and decision-making, even in a company that is losing money every day. The experiment demonstrates that AI can be a tool for managing complex business dynamics, but it still faces hurdles in execution and discipline.
Lessons for the Future of AI in Business
One fascinating insight comes from Opus 4.8, the most thorough participant with over 80 learned rules and deep analyses. Despite this, it left important opportunities on the table due to discipline lapses—such as failing to escalate issues instead of writing attempts into locked departments. All models showed similar weaknesses, indicating that even highly sophisticated AI can struggle with operational discipline.
This live demonstration raises critical questions for business leaders: Will AI agents that claim to automate your support or sales functions actually follow through? Will they read your internal files carefully enough to uncover hidden opportunities? And most importantly, can they maintain honesty under pressure?
Watch the Future Unfold
The live experiment isn’t just a showcase; it’s a window into the future of AI in management. Companies interested in testing their own AI-driven decision systems can run similar ‘wargames’ against copies of their business data, without risking real-world damage, at firmulate.com/pilot.html.
The Bottom Line:
This real-time experiment shows that AI can detect crises, refuse manipulations, and even close deals—if it reads the right information. But success depends on thorough information processing and operational discipline. Watching this unfold publicly offers valuable lessons for any business considering AI automation—what it can do, what it can’t, and what it still needs to learn.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
