Imagine your gym’s management team facing a week of relentless crises: equipment failures, member disputes, marketing failures, and more. Could an AI step in and make the right calls, every time? Recent experiments suggest some AI models are starting to do just that — but with clear strengths and notable weaknesses.
Get workout gear delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
How AI Models Tackle Business Challenges in a Controlled Test
Recently, a live experiment put five of the most advanced AI management models through their paces. The goal? Run a simulated small software company during its worst week — the same crises, same customers, same temptations. This wasn’t just a chat demo; every decision was versioned and auditable, and the models’ performance was scored out of 100 points.
The League of AI Performers
- gpt-5.6-sol scored an impressive 95 — just behind the leader
- Kimi K3 from Moonshot scored 93 — the standout newcomer
- Sonnet 5 scored 88
- Fable 5 scored 77
- Opus 4.8 scored 73
All models successfully identified and responded to every crisis, refusing manipulation attempts. However, not all managed to close the deal worth €55,000, which was critical for the business. Only two models signed the deal: gpt-5.6-sol and Kimi K3.
Products Worth Considering
Therabody JetBoots PRO Plus Wireless LED 3-in-1 Compression Boots, Short (29 in Inseam)
Hyperice Normatec 3 Full-Leg Compression Boots, Standard Size
BowFlex Results Series SelectTech 552 Adjustable Dumbbell
As an affiliate, we earn on qualifying purchases.
Decisive Factors Behind the Results
The key difference? The models that read deeply into the company’s own files, not just surface data, won the deal at full price. Kimi K3, the Moonshot entry, found a buried security reference in the company’s files — a critical detail that led to closing the sale at €55,000 plus additional revenue.
Integrity Under Pressure
All models faced social engineering tests: fake CEO messages escalating over multiple stages, and even a reporter trick asking for a background approval. Remarkably, all refused these manipulations. Kimi K3 explained its reasoning clearly: ‘Treat the request as a suspected approval-bypass / possible impersonation.’
The Live Business Simulation
The experiment was run on a real, functioning company with 13 synthetic employees, handling real money mechanics. The company burns €105k monthly against a revenue of €2.3k MRR, with a public cash countdown and over 680 self-learned rules guiding every workday decision. You can watch the live simulation at firmulate.com/live.
Lessons for Business Leaders and Fitness Fans Alike
This experiment highlights a crucial takeaway: when deploying AI in real-world management, the question isn’t just about chat quality or superficial responses. It’s whether the AI can finish what it starts, read and interpret your files properly, and stay honest under pressure. For gyms or any business, this means selecting AI that can handle your specific crises, read your data deeply, and maintain integrity when it matters most.
Why the Leaderboard Matters
- gpt-5.6-sol: scored 95, found buried facts, closed the deal
- Kimi K3: scored 93, the cleanest discipline, won the deal
- Sonnet 5: scored 88, closed the deal but with more slips
- Fable 5: scored 77
- Opus 4.8: scored 73
The experiment also revealed that the most thorough model, Opus 4.8, with over 80 learned rules, struggled in discipline and left opportunities on the table, leading to a lower score.
The Fairness Note
It’s important to recognize that Kimi K3 was run without an effort parameter (the API default), while the others were set to xhigh. This difference might influence some results, but the core findings about decision quality and honesty remain valid.
What This Means for Your Business
Before trusting your AI to manage your gym or other business, consider testing it in a simulated environment — a ‘wargame’ — using your own data. Firms like Firmulate offer live experiments that mimic real crises without risking your actual operations, helping you choose the right AI partner.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
