Europe's AI Divide Is Already an Enlargement Problem

URL SCAN: Europe's AI Divide Is Already an Enlargement Problem
FIRST LINE: Montenegro’s business register counted 63,823 active business entities in 2025.

THE DISSECTION

This is a competent policy-diffusion memo with a strategically inadequate diagnosis. It correctly identifies Montenegro’s micro-firm economy, the statistical blind spot below ten employees, and the concentration of AI gains in firms able to finance data, software, and training.

Its operative move is to turn AI into a delivery problem: measure participation, distribute grants, provide training, build hubs, and monitor adoption. That is administratively coherent. It is not a solution to the underlying discontinuity. The text assumes the objective is to spread AI productivity rather than determine whether human labor remains economically necessary after AI deployment.

THE CORE FALLACY

The text confuses diffusion with democratization. Giving micro firms access to AI does not give them durable economic sovereignty. It may instead expose them to stronger competition from larger firms that possess superior data, capital, integration capacity, and bargaining power.

The EIB’s four per cent productivity gain is an early, conditional measurement—not evidence that the employment circuit survives. Under the Discontinuity Thesis, productivity gains are the leading edge of labor displacement. Once cognitive automation becomes cheaper and more capable than human labor, wider adoption accelerates the destruction of economically necessary work. A grant can help a firm automate; it cannot guarantee that the workers displaced by that automation retain purchasing power or productive leverage.

The article treats continued use after twelve months as evidence of success. Under the harder test, the question is whether the firm remains competitive without requiring the same human labor. Adoption persistence may measure the persistence of the transition, not convergence.

HIDDEN ASSUMPTIONS

• Micro firms can overcome structural scale disadvantages through advice, subsidies, and shared services.
• Productivity gains will translate into wages, employment, or broad local prosperity.
• Single-market integration will distribute technological gains instead of selecting for concentration.
• Training can preserve worker value after AI reduces the amount of necessary cognitive labor.
• Institutional capacity can keep pace with competitive automation.
• Verified technology uptake is a meaningful proxy for economic viability.
• The problem is unequal access to AI, rather than the shrinking need for human participation itself.

These assumptions convert a capital-ownership problem into a programme-design problem. That is the bureaucratic alchemy at the center of the piece.

SOCIAL FUNCTION

This is partial truth functioning as transition management and ideological anesthetic. The delivery failures are real: small firms miss programmes, cannot absorb fixed costs, and lack the staff to implement complex systems. The article is useful on that narrow mechanism.

But it redirects attention from the terminal question—who owns and controls the productive systems—to whether every shop, hotel, and professional service firm has received adequate technical assistance. It offers administrators measurable milestones while leaving the ownership structure and employment collapse untouched. The result is a cleaner dashboard over a worsening battlefield.

THE VERDICT

The article accurately describes how AI gains will initially bypass Montenegro’s micro firms. It fails to grasp that universal diffusion would not restore the post-war employment–wage–consumption circuit; it would extend automation into the firms currently excluded from it.

Enlargement funding may delay divergence and create temporary transition niches. It cannot make mass human labor indispensable once P1, P2, and P3 are in force. The proposed policy does not defeat obsolescence. It distributes the accelerant and labels the distribution convergence.