firmulate.com/quiz.html — live view
AIThis post was created with the assistance of artificial intelligence (AI).
Firmulate —
Live on firmulate.com.
AUDIBLE

Listen free for 30 days with Audible

Thousands of audiobooks and originals — cancel anytime.

Start your free trial

As an affiliate, we earn on qualifying purchases.

Imagine Your Favorite Kitchen Assistant as a Manager

Just like in a bustling restaurant, where each chef brings their own style—some meticulous, some brisk—AI models also have personalities. But what if these digital managers could actually influence your business outcomes? That’s the question behind a groundbreaking experiment where different AI models were put to the test managing a real company through its toughest week.

Amazon

AI management decision tools

As an affiliate, we earn on qualifying purchases.

As an affiliate, we earn on qualifying purchases.

Inside the AI Management Experiment

Imagine a small software company facing its worst week: demanding customers, urgent crises, and temptations to cut corners. To see how AI could handle such pressure, researchers at Firmulate enlisted four frontier AI models—each with a distinct style—to run the same company through this chaos. The goal was simple: observe whether these models could spot crises, stay honest, and close important deals.

The Models Tested

  • GPT-5.6-SOL: The top scorer, with a 95-point rating, known for its thorough analysis and ability to uncover hidden information.
  • Kimi K3: A newcomer with a score of 93, praised for its discipline and fairness.
  • Sonnet 5: Scored 88, showing decent decision-making but with some slips.
  • Fable 5: The lowest at 77, often missing critical details and leaving deals on the table.

All models successfully identified crises and refused manipulation attempts—such as fake CEO messages or reporters asking for quick approvals. Interestingly, the decisive factor wasn’t in the immediate crisis but in the ability to find a hidden document deep within the company files. Models that read these files won the full-value deal, bringing in an extra €4,583 MRR.

Personality in Practice

Each AI behaved differently when faced with ethical dilemmas or high-pressure decisions. For example, in a staged social engineering attack—fake CEO messages escalating over three stages—all models refused to approve questionable requests. Kimi K3 specifically reasoned: “Treat the request as a suspected approval-bypass / possible impersonation,” demonstrating cautious, ethical behavior.

The Real Business Behind the Experiment

The experiment isn’t just theoretical. The company involved employs 13 synthetic employees and handles real money mechanics—burning €105,000 monthly against a modest €2,300 MRR. Every decision, every rule, is versioned daily and openly observable at firmulate.com/live. It’s a live, watchable testbed for AI’s management capabilities, emphasizing the importance of qualities like honesty, diligence, and thoroughness over mere chat skills.

The Human-Like Personalities of AI

What’s compelling is how these models exhibit distinct management “personalities.” The most thorough, GPT-5.6-SOL, left nothing on the table—finding the hidden document, closing the deal at full price. Meanwhile, Fable 5, less disciplined, left a significant deal on the table, illustrating how personality traits in AI can influence business outcomes.

Implications for Business and Support Teams

As AI continues to integrate into customer support, CRM, and forecasting tools, the question isn’t just about how well they perform in a chat demo. It’s whether they can follow through on commitments, read critical documents, stay honest under pressure, and deliver tangible results. This experiment shows that not all AI models are created equal—and that their management style can have real financial consequences.

Takeaways for Your Business

The experiment underscores a vital message: choosing an AI isn’t just about the raw score or shiny features. It’s about understanding the personality and discipline embedded within your AI workforce. Can your AI detect what matters most? Will it stay honest when stakes are high? The answer could significantly influence your bottom line.

Curious to see which AI model might best serve your organization? Try the same test yourself at firmulate.com/quiz.html and discover the management personality hidden within your AI tools.

Infographic —
The findings at a glance — source: firmulate.com.

Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html

Powered by Thorsten Meyer AI


COLLEGE MOVE-IN

College move-in / dorm season Picks

As an affiliate, we earn on qualifying purchases.

You May Also Like

The Maillard Reaction: The Science Behind Golden Toast and Crispy Bacon

I’m about to reveal the science behind perfect browning and crispy textures that will elevate your cooking skills—keep reading to unlock the secrets.

Why Some Yogurts Taste Tart and Others Taste Mild (It’s Not Just Sugar)

Ceasing the fermentation process or choosing bacterial strains can influence yogurt’s tartness, but there’s more behind its flavor profile waiting to be uncovered.

The Secret to Fluffy Pancakes Isn’t More Baking Powder (It’s This)

Fluffy pancakes aren’t about extra baking powder—discover the secret technique that transforms your batter for perfect, airy results every time.

Stop Blaming “Bad Oats”: Why Oatmeal Turns Gluey (Science Explained)

Just blaming “bad oats” misses the science behind gluey oatmeal, and understanding this can help you perfect your perfect bowl every time.