Crystal ball for everyone: what happens when every company is managed by the same AI?

Managing a company has always been the art of making decisions under uncertainty. Which product to launch? What price to set? When to offer a discount? How much to order for the season? Which channels to expand? Each of these decisions required experience, intuition, and — let's be honest — luck.

Now imagine that all these decisions can be automated. The industry knowledge base is public. Mental models — defined and open. AI knows how your industry's customer thinks, what motivates them, what they fear. All you need to do is plug your data (sales, inventory, customer behavior) into a publicly available model — and you get an answer. Not after a month of analysis. In a second.

Sounds like science fiction? It's already happening.

Case study: children's product retail

Let's take a concrete example. A retail company selling children's clothing, toys, and accessories. A brutal industry: low margins, extreme seasonality, emotional and demanding customers. Every decision matters.

For years, such companies built their advantage on tacit knowledge — on the experience of buyers, the intuition of category managers, supplier relationships, the ability to read the market. That knowledge was their capital. Their IP, though nobody called it that.

Now imagine someone formalized that knowledge. Created a 360° map of mental models — a complete grid of all key beliefs, rules, and mental shortcuts that steer decisions in the company. And published it. For free.

The 360° map: a retail company's mental models

Here's what such a map looks like for a company in the children's products segment. 37 mental models, 6 areas, full decision spectrum:

This isn't abstract theory. It's a complete decision map for a company. Each of these models is know-how — knowing how the customer thinks, reacts, what they avoid. For decades this knowledge was protected in managers' heads, in organizational culture, in informal rules. It was hard to copy.

But what if someone formalizes it and makes it public?

Thought experiment: a crystal ball for everyone

Let's assume the mental model map above is publicly available. Every company in the industry can use it. Now let's add AI:

1. Step 1: Company A takes the 37-model map and feeds it as a decision framework to a language model.
2. Step 2: It pastes in its data: 12 months of sales, inventory levels, margins by category, return rates, website traffic data.
3. Step 3: AI analyzes the data through the lens of mental models and generates recommendations: "Increase accessories by 40%, reduce inventory in sizes 68–74 after March, shift marketing budget to parent-generated content, avoid aggressive promotions on premium products."
4. Step 4: The company implements the recommendations. Immediately.

Brilliant. Competitive advantage. For three weeks.

Because Company B did exactly the same thing. And Company C. And D, E, F, G, and the other 200 companies in the industry. Everyone has the same crystal ball. And everyone is looking into it simultaneously.

When everyone has an advantage, nobody does. A crystal ball that sees the future — but the same future for everyone — is useless.

The clone paradox: who does AI really help?

This isn't a hypothetical problem. It's happening right now, before our eyes, in every industry.

E-commerce: the testing ground

Online retail is the perfect example, because here replicating someone else's functionality takes minimal time.

• Store A introduces "similar product recommendations" — within a week everyone has them.
• Store B adds "buy now, pay later" through Klarna integration — within two weeks the entire industry does it.
• Store C personalizes the homepage based on browsing history — within a month everyone on Shopify does it too.
• Store D launches an AI chatbot for customer service — the same API is available to anyone for $20/month.

In traditional retail, advantages were built over years: location, supplier relationships, brand recognition. In e-commerce, every innovation is copied within weeks. Building advantages loses its meaning because before you can measure the effect, the competitor already has it.

Now add AI to the equation

Take a company selling children's products online. It has 5,000 SKUs, 20,000 orders per month, 3 years of historical data. It feeds everything into AI with the mental model framework. AI says:

• "Parents buy ahead of time" → set up layette bundles with a 12% discount
• "Safety > price" → highlight safety certificates on product cards
• "Cash flow is king" → limit promotions on slow-moving products, liquidate for cash instead
• "The margin is in accessories" → cross-sell socks, hats, bibs with every clothing purchase
• "Seasonality is brutal" → order the autumn collection in March, not May

These are good decisions. Smart. Based on real industry knowledge. But these are exactly the same decisions that anyone who uses the same framework with the same models will make. And everyone will use it.

The race to convergence

Economists have a term for this: strategic convergence. When all companies in an industry respond to the same signals with the same tools, their strategies become identical. Prices, assortment, marketing, logistics — everything converges to a single point.

Under normal conditions, convergence is a slow process. Companies learn from each other over years. They copy solutions quarter by quarter. There's time for feedback loops.

But AI shortens this process to days. When every company can in a single afternoon:

1. Download a public industry mental model framework
2. Feed it their own data
3. Receive identical recommendations to the competition
4. Implement them immediately

— then convergence isn't a process. It's an event. A quantum leap after which all companies stand at the same point.

In a world where every company has access to the same knowledge, the same tools, and the same models, the only thing that differentiates them is their input data. And input data — is just a matter of time.

What really remains when knowledge is free?

If mental models are public. If AI is available to everyone. If a competitor's data can be approximated within a week. What truly differentiates one company from another?

The answer is unsettling: almost nothing.

In theory it could be:

• Speed of implementation — whoever implements AI's recommendations first has a short-lived advantage. But "short-lived" here means hours, not months.
• Unique data — data about your customers' behavior that the competitor doesn't have. But in e-commerce, data leaks through cookies, pixels, scrapers, and affiliates. Data protection as an advantage is an illusion.
• Brand and trust — the only thing that can't be generated by an algorithm. But it takes years to build and a single scandal to destroy.
• People — but if AI makes decisions better than people, what is the role of people?

The endgame scenario: a market of clones

Let's look five years into the future. In the children's products industry in Poland, 300 online stores operate. Each uses a publicly available mental model framework. Each feeds data into the same or very similar AI. The result:

• Prices are identical (AI determined the optimal point).
• Assortment is identical (AI selected the same bestsellers).
• Marketing looks the same (AI generated the same copy based on the same models: "warm colors, photos of children, social proof from parents").
• Logistics identical (AI optimized the same supply chain).
• Customer service identical (the same chatbot, the same scripts).

300 stores. One store multiplied by 300. A market of clones.

What then? In a market of identical players, those who win are those with capital to survive a price war. Those who can operate on smaller margins for longer. Those who can burn cash on customer acquisition and wait for the smaller ones to drop out.

Meaning — again — the biggest wins. Not the smartest. Not the most innovative. The biggest.

The fundamental question

And so we arrive at the question that should keep us up at night:

Does a world where knowledge is free and instantly accessible inevitably lead to monopolies? Does the democratization of management tools paradoxically mean the end of competition?

Because if so — we have a problem. Not a technological problem. A civilizational one.

The entire market economy rests on the assumption that companies compete because they have different ideas about how to satisfy customer needs. That diversity of strategies creates diversity of offerings. That competition drives innovation.

But when all companies think the same — because the same AI thinks for them — competition becomes a fiction. Not a rivalry of ideas, but a race of capital. And the capital race is always won by the one who has more.

We're not writing this to demonize AI. AI is a tool — powerful, useful, transformative. The problem doesn't lie in the tool. It lies in what happens when everyone uses the same tool in the same way.

Mental models — like these 37 points in our 360° map — are the quintessence of industry wisdom. Publishing them is an act of opening. An act of No IP in its purest form. But the consequences of that opening aren't simple and aren't unambiguously good.

Perhaps it will turn out that the real competitive advantage of the future won't lie in knowing what to do — but in the courage to do something different from what AI recommends. Paradoxically, in a world of omniscient algorithms, the greatest value may lie in human irrationality.

Because the only thing AI can't copy is the idea of not listening to AI.