The pottery magic: How AI is reframing the competition in business

The main thing AI has done for business is fundamentally change competition. When AI is in skilled hands, it has effectively reduced the cost of feedback, which is essential for building better products, to almost zero.

The rise of AI closely resembles the famous “Pottery Experiment”, an anecdote illustrating the power of quantity over perfection, popularised in James Clear’s book Atomic Habits. A ceramics professor reportedly divided students into two groups. The first group was graded solely on the quality of a single, perfect pot produced over the semester. The second group was graded purely on the quantity, measured by weight, of pots they completed. Surprisingly, the students who focused on quantity produced the highest-quality work because they learned through rapid iteration and mistakes, while the “quality” group wasted time theorising.

Before AI, companies highly valued resources and well-structured processes to deliver a product. These elements kept the business afloat, enabled growth, supported new product launches, and helped attract customers.

Now, the game has fundamentally pivoted. AI is not just a feature. It delivers hyper-fast market feedback, which has become the ultimate moat. That demands organisational flexibility, non-obvious thinking, and fidelity to genuine product-market fit.

Processes and resources as a burden

Processes and resources that were once essential for growth are increasingly becoming a burden. Corporations necessarily involve bureaucracy, with processes designed to preserve resources and therefore growth. These processes require multiple layers of approval, with larger decisions involving more people. This no longer makes sense. The resources required to fund product development are no longer a competitive advantage, because intelligently applied AI can deliver comparable results at a fraction of the cost in money and human hours. Crucially, these decisions can now be made swiftly by a single founder.

Can you meaningfully compare an AI-developed product with one built by a large professional department and tested by another? There may be nuances and differences, but the person paying for the product is unlikely to notice a substantial gap in how well it satisfies their needs. In many cases, the quality of products built with AI tools may even be higher. The key point for competition is this: AI neutralises the advantage of resources and shifts attention to the second advantage, flexibility versus process rigidity.

When a large corporation identifies a market opportunity and decides to launch a new product, it faces a bureaucratic journey that can last months. Processes are designed to manage resources carefully, and quick results are unrealistic when the human hours of development, testing, and multiple departments are at stake.

To meet market demand, corporations often have to fight their own rules. Their processes do not allow them to act in ways that are not formally documented.

A startup reacts very differently. Using AI, it can launch a product and test a market hypothesis as quickly as possible. It is not constrained by formalities or rigid processes. The focus is on learning fast whether people need the product, and if they do, what needs to change to make it better. This is where a small startup concentrates its resources: achieving product-market fit. In extreme cases, highly successful projects consist of just one person who is exceptionally skilled at managing AI.

A third drawback of extensive resources is that they consume resources. Departments require salaries, which are reflected in the final product price. A startup using AI can significantly reduce costs, offer lower prices to customers, attract more users, and continuously improve the product based on feedback.

The pottery effect for startups

If an AI-enabled startup and a corporation identify the same market opportunity at the same time, the outcome after a year is revealing. The corporation may have tested two hypotheses using a hundred engineers, while the startup may have tested a hundred hypotheses with just two engineers. The result is that the startup’s product aligns far more closely with market demand, and in some cases even creates demand that did not previously exist. This advantage comes from exponentially greater feedback volume and the flexibility to respond to it.

The natural limits of this model are capital-intensive industries such as space, defence, and medicine, where the cost of experimentation is too high to allow fast and cheap customer feedback. The same applies to B2G sales and other capital-intensive sectors.

How corporations should adapt

For large corporations to compete with fast-moving, AI-backed rivals, they must emulate their approach, not the other way around. First, they must recognise that slow, traditional bureaucratic structures are a major obstacle. The corporate model must evolve to prioritise speed and experimentation.

A critical step is establishing autonomous AI squads operating within governed sandboxes, meaning controlled environments with simplified compliance requirements. These teams must be fully disconnected from standard bureaucratic decision-making processes. This enables rapid iteration and deployment of AI prototypes without long approval cycles.

New key performance indicators should measure learning velocity and the speed of technology implementation, signalling a clear shift away from legacy metrics tied to slow, long-cycle projects. Corporations should also create a comprehensive catalogue of internal data to ensure easy access for model training.

The nature of AI-powered rivalry

The new rivalry model will not immediately collapse traditional businesses or halt their growth. The larger the organisation, the slower the decline. That is precisely the problem. Corporations fail to adapt because they do not perceive an immediate threat, as annual reports still show positive growth. Growth slows gradually, first by fractions of a per cent, then by whole percentages. Meanwhile, new businesses may grow by thousands of per cent per year, eventually challenging incumbents that are entirely unprepared to respond.

This dynamic is well illustrated by the relationship between Intel and Nvidia, even though AI experimentation was not a factor at the time. “If you don’t have a microprocessor, what else do you have to sell?” Intel CEO Paul Otellini said in 2009, dismissing Nvidia’s claim that the industry was shifting towards graphics chips. Nvidia’s success was fuelled by Intel’s dominance, which allowed its rival to grow quietly and ultimately overtake it a decade later.

With the added force of AI, such shifts in dominance will happen much faster.

The post The pottery magic: How AI is reframing the competition in business appeared first on EU-Startups.