Imagine running your outdoor retail business through its toughest week—crises, manipulative customers, and tough negotiations—while an AI monitors and manages every step. Would it outperform human managers? Recent experiments with cutting-edge AI models suggest it might, revealing surprising strengths and weaknesses in how these digital managers handle real-world pressures.
The Live Business Wargame: Putting AI to the Test
In a groundbreaking live experiment, four state-of-the-art AI models were tasked with managing a small software company during its worst week. This wasnt a mere simulation; it was a fully functioning, real-time business environment with actual money mechanics—monthly burn of €105,000 against a revenue of just €2,300, plus a countdown of public cash. Every decision made by these models was real, auditable, and focused on navigating crises, customer manipulations, and internal dilemmas.
How the Experiment Worked
Each AI model faced the same scenarios: identical customer complaints, crises, and temptations. The models’ decision-making was monitored, with their actions and reasoning openly recorded. Their goal? Survive the week, close deals, and uphold integrity in a high-pressure environment. The models were ranked based on their overall performance, including their ability to identify hidden information, remain honest, and ultimately close a crucial deal worth over €4,580 in monthly recurring revenue (MRR).
AI business decision making software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Results: The Good, the Bad, and the Surprising
All four models successfully identified and responded to every crisis—no missed alarms there. They refused every manipulation attempt, including a staged social engineering attack involving fake CEO messages and a reporter trick, with all models refusing to be duped. Notably, Kimi K3’s reasoning was explicit: ‘Treat the request as a suspected approval-bypass / possible impersonation.’
Where differences emerged was in deal closure. Only two models signed the €55,000 deal their own analysis had earned. The other two, despite diagnosing the situation correctly and presenting solid pitches, left the deal on the table, citing discipline slips and unexecuted follow-through.
The Hidden Weakness: Reading the Files
Digging deeper, the decisive edge came from reading company documents. The models that examined internal files uncovered a critical piece of information buried two documents deep—information that led to closing the full-price deal. Conversely, models that didn’t delve into these documents missed the opportunity, costing the company €4,583 in potential monthly revenue.
Personality Profiles of the AI Managers
The models displayed distinct ‘management personalities’:
- gpt-5.6-sol: Most thorough, identifying the hidden fact and closing the deal—top score of 95.
- Kimi K3: The newcomer with a disciplined approach, closing the deal at 93.
- Sonnet 5: Slightly more process slips, finishing at 88.
- Fable 5: Similar to Sonnet, with more slips, ending at 77.
The most comprehensive model, Opus 4.8, with over 80 learned rules, was surprisingly the lowest performer in deal closure—demonstrating that depth of analysis doesn’t always translate into execution. Interestingly, Kimi K3 operated without the default effort parameter, possibly contributing to its clean performance.
Implications for the Future of Business Management
This experiment isn’t just about AI beating or matching humans in a game; it raises fundamental questions about automation in real-world business decisions. The models’ ability to detect hidden information, refuse manipulative tactics, and stay disciplined under pressure are critical factors that could revolutionize how companies manage risks and opportunities.
For outdoor and garden retail companies considering AI solutions, these findings highlight the importance of testing AI in realistic, high-stakes scenarios before deployment. A tool that can identify critical internal data and remain honest under pressure isn’t just a nice-to-have—it’s essential.
Try It Yourself
Curious whether your own business could benefit from this kind of AI management? You can run your own wargame against a read-only export of your operations. It’s a safe, no-risk way to assess how AI might perform in your unique environment. More details are available at firmulate.com/pilot.html.
Why It Matters
In a world where AI might soon oversee customer relations, supply chains, or strategic decisions, the question isn’t just about how well it communicates but whether it can stay honest, read critical internal information, and close deals reliably. This live experiment demonstrates that AI can do all of these—if tested and calibrated carefully.
Real AI management tests reveal that while all models detect crises and refuse manipulations, only some can read critical internal info and close deals reliably. The future of AI in business depends on rigorous real-world testing—are you ready to wargame your AI workforce?
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
