Working Principles

Julien Thibaut — Brussels, May 2026. A set of stress-tested principles, not a concluded thesis. To be argued with, extended, and connected.

I. On AI as Infrastructure

1. AI amplifies existing competence — it does not replace the need for it.
AI is a multiplier on human quality. A weak operator gets weak results faster. A strong operator gets strong results dramatically faster. The gap between a skilled and an unskilled user of these tools is widening, not closing. Expertise, craft, and judgment are more valuable in an AI-assisted world, not less.

3. Vibe coding without specification is a trap.
Natural language prompting works for prototypes and exploration. The moment you build something that must be maintained, debugged, or extended, you need written specifications as the source of truth — not the code the AI generated. The people having real success with AI coding are not "vibing." They are specifying: writing intent first, then generating. The discipline is pre-AI. The speed is new.

5. Build businesses that use AI as cheap tools — not as infrastructure or product.
Building a business whose operations run on AI at scale is a bad bet.
Imagine this venture :

Building a POS system for Belgian cheese shops — bilingual, PEPPOL-ready, cheese-platter aware — with three people and 100,000 euros in a year. that's cheap to build and you will be able to sell at much lower prices that currents products. niche, cheap, useful, in demand . great ! what a time to live in !

But here is what is also true: by the time you visit cheeses shops a year later with you new, cheap, amazing POS, others have finished, are finishing or will build soon competitive products. Your product is excellent by the standards of a year ago. It may not be excellent by the standards of today — and almost certainly will not be by the standards of next year. what a time to live in...

If AIs are the new, mass produced, golden geese - then the business of selling golden eggs will become a commodity business. Prices compress. Margins compress. The advantage evaporates before it compounds.

This is not a new pattern. The investors who built Britain's railways went broke in 1847. Gutenberg died with nothing. McLean's ship containers company failed twice. In every case, the value did not go to the inventors of the enabling technology. It went to the smalls, medium businesses who reorganised their operations around the new abundance. ( it also did not last, as they also got price compressions and margins compressions over time. but there was genuily money to be made and opportunities to starts. )

The EU's "lag" in the AI race is irrelevant . You do not need to breed your own geeses. nor to sell golden eggs.

The question is not who invented the goose. who sells the best golden eggs
The question is: what are you going to make with all these eggs?

GPT model : A new technology gets invented. The inventors (most likely a lucky iterator with a beter business model than the inventor himself ) profit first. Then the people who build products on top of it profit. Then businesses that use those products get an edge over competitors who don't. Then even the laggards adopt it, the edge disappears, and everyone just has lower costs. Then consumers benefit from those lower prices. Then even that comperativly equalizes.

At each step, early movers win briefely if not at all.

THE TRUE WINNERS at each layer is not just whoever adopts first — it's whoever builds the hardest-to-replicate complementary assets during their window.

the window is NOT about using AI faster than competitors. It's about using the window to build something that doesn't depend on AI remaining an advantage. Trust, brand, knowhows, culture, innovative mindsets— these are the kinds of complementary assets that survive commoditization.
exemples :

• Distribution & relationships — Coke owns bottling franchises; the tech is free to copy, the shelf space is not
• Operational culture — Toyota published TPS freely for 40 years; GM still couldn't replicate it
• Tacit industry know-how — IBM pivots every decade because it knows how enterprises buy, fear, and decide
• Brand as trust/accountability guarantee — Red Hat sells free Linux for $34B because it sells certified reliability
• Proprietary data flywheel — Walmart's Retail Link makes suppliers restructure around Walmart's data, not vice versa
• Regulatory position — PEPPOL integrators who embedded themselves before the January 2026 mandate own the accountant's recommendatio
• Switching costs baked into workflows — SAP's ERP is not the best software; it is the software your finance team has been trained on for 10 years, your audit trail is inside, and migrating costs more than the pain of staying
• Network effects — Visa and Mastercard are not payment technology companies; they are trust networks where each new merchant and cardholder makes the whole more valuable to everyone else, making a newcomer's better tech irrelevant at launch
• Intangible assets: certification & standards ownership — Dolby does not make speakers or microphones; it owns the audio certification standard that hardware makers must licence, making it structurally embedded in every device regardless of who built it
• Community & ecosystem lock-in — Salesforce's real moat is not its CRM features (Hubspot matches them); it is 170,000+ third-party apps built on its platform by developers who are not going to rebuild them for a competitor
• Reputation in high-stakes decisions — McKinsey charges 10–20x what comparable analysts charge, not for better Excel models, but because recommending McKinsey protects the executive who hired them; the brand absorbs career risk, and that is worth any price differential
  
• Factory floor redesign know-how (electricity era) — Factories that restructured their entire floor plan around decentralised electric motors in the 1920s did not win because they had electricity. Every factory eventually had electricity. They won because the redesign created organisational habits, supply chain coordination, and quality consistency that competitors had to rebuild from scratch. The asset was the new way of working, not the power supply.
• Land and access to the new network (railways) — When British railways expanded in the 1840s, the investors in the railway companies mostly went broke. The landlords and merchants whose towns were connected to the network captured the value — in the form of rising land prices, new markets, and lower freight costs — without having built a single mile of track. The complementary asset was proximity to the infrastructure, not ownership of it.
• Coal and steel supply chains (steam engine era) — James Watt improved the steam engine but made most of his money defending patents, not from the industrial revolution he triggered. The fortunes were built by the coal mine owners and ironmasters who supplied the inputs the engine needed and who had the operational scale to deploy engines across entire industries. The asset was vertical integration into the supply chain of the new technology, not the technology itself.
• Publishing networks and literacy infrastructure (printing press) — Gutenberg died broke. The wealth flowed to the Frankfurt Book Fair merchants, the university press networks, and eventually the Protestant church — institutions that built distribution, editorial curation, and trusted authority in a world suddenly flooded with cheap text. The complementary asset was the ability to select, certify, and distribute in a world of information abundance — exactly the problem AI creates today.
• Embedded work organisation (IT/electricity parallel) — MIT Sloan's research on IT adoption showed that firms invest 10 times more in intangible organisational assets associated with new technology than in the technology itself — and that firms investing heavily in both IT and work reorganisation have measurably higher market valuations than those who invest in IT alone. The asset is not the tool; it is the redesigned way people use the tool together.
• Geographic concentration and talent clusters (industrial revolution) — Manchester, Sheffield, and the Ruhr Valley did not win the industrial revolution because they had better machines. They won because they concentrated the engineers, the suppliers, the financiers, the skilled workers, and the informal knowledge networks that made deploying new technology faster and cheaper than anywhere else. The complementary asset was the ecosystem itself — which is why Silicon Valley still matters even when compute is global.

II. On Europe's Actual Economy

6. Europe's productivity "gap" versus the US is concentrated almost entirely in the digital tech sector — not the broader economy.
FEPS (2025), citing Darvas (2023), shows that in per-capita output and output per hour worked, the EU has not fallen significantly behind the US. The gap is driven by US tech superstars — IT services and computer/electronics manufacturing — and partly by COVID-era US fiscal stimulus that boosted measured productivity via unemployment spikes rather than genuine output gains. ILO data puts Western European labor productivity per hour at $83 vs. $81.80 for the US. Draghi's own report confirms the sectoral concentration: IT services and computer manufacturing were the main contributors to US productivity over the last two decades, while the rest of US industry stagnated. Honest caveat: the OECD and AMECO data do show a real US lead that has diverged since the 2010s. The gap is real but narrow and sector-specific — not pervasive.

7. Europe runs on leisure, culture, and discretionary premium spending — not on scale.
Tourism alone represents 10% of EU GDP and nearly 2 million SMEs. The EU cultural sector comprises 2.03 million enterprises generating €199 billion in value added — 6.3% of the commercial economy. Government cultural expenditure is €81.1 billion annually. The dominant economic activity in Europe is embedded local service, cultural consumption, premium craftsmanship, and experience — not platform businesses or high-growth startups. Honest caveat: the Draghi report does not validate this framing — it focuses on technology scale and industrial competitiveness. Both are simultaneously true: Europe runs on local service economically, and Europe's policy elite believes it must compete in advanced technology strategically. These are not the same thesis.

8. The 50+ consumer is the dominant spending cohort in Europe and is systematically underserved by innovation.
The 60+ cohort represents 28% of Europe's $13 trillion total spending power today, rising to a third of $16 trillion by 2030. In France, people aged 50+ hold nearly half of total national purchasing power and provide 45% of total demand. Germany's silver purchasing power alone is €316 billion. People over 50 generate 34% of global GDP and hold nearly half of European household wealth. VC-backed startups systematically optimize for under-35 users — the gap between where the money is and where innovation is pointed is structural, not incidental.

III. On Owner-Operators and Service Quality

9. Owner-operated businesses structurally deliver better service than managed ones.
Over a century of organizational research, spanning more than 100 studies across multiple economies, shows that employee and owner-operated structures produce better productivity, higher job quality, stronger firm survival, and more committed teams compared to purely wage-labor equivalents. The mechanism is alignment of incentives with outcomes.

10. Ownership raises service quality at the point of human contact.
The customer-facing difference between an owner-operated business and a managed one is not a matter of training or policy — it is structural. Owner-operators invest in long-term customer relationships because those relationships are their livelihood, identity, and social standing. They take initiative that salaried employees have no incentive to take. For premium consumers who are highly sensitive to human interaction quality, this is the deciding differentiator.

11. Young European talent is not unmotivated — it is mis-incentivized.
The European brain drain is not primarily wage-driven. It is ownership-driven. Skilled young workers leave for contexts where they can have a stake, make real decisions, and see the connection between their effort and its outcome. This is a structural failure of European corporate and institutional culture to offer ownership at the right career stage — not a cultural deficiency of young Europeans.

12. Moral direction accelerates commitment in teams — but only when structurally embedded.
Bregman's Moral Ambition (2025) documents what high-performing teams share: a traceable connection between daily work and a problem worth solving. This is not about mission statements. It functions when it is embedded in the mechanics of the business — who owns it, how decisions are made, what it refuses to optimize for. Purpose as branding does not produce commitment. Purpose as structure does.

IV. On the Draghi Signal

16. The Draghi report is the most influential European policy document in a generation — and it signals that Europe is seriously trying to define its own economic trajectory.
Published September 2024, commissioned by the European Commission, endorsed by every EU member state and the European Parliament. Von der Leyen described it as the foundation of the Commission's political guidelines for 2025–2030. Its core diagnosis: Europe faces "slow agony" if it does not close its innovation gap, reduce energy dependency, and raise investment by €750–800 billion annually. Three things in it are directly relevant here: (1) it calls for reducing SME regulatory compliance costs by up to 50%, explicitly identifying bureaucratic burden as a core competitiveness problem; (2) it frames digital and energy sovereignty as existential priorities, which directly supports the open-model / local-energy AI direction in Principle 5; (3) its intellectual weight signals that Europe's most serious thinkers have stopped treating the US model as the destination and are now trying to define a distinctly European economic path. What it does not validate: it does not frame leisure, culture, or premium local service as strategic priorities — its instinct is toward consolidation, large-scale industrial champions, and advanced technology. This is a genuine tension with Principle 7, not a contradiction to be smoothed over.

V. On the Civil Economy Frame

13. The market has a pre-capitalist tradition built on reciprocity, not just exchange.
Economists Bruni and Zamagni document a tradition — the civil economy — rooted in Italian Civic Humanism, which places reciprocity and the common good as constitutive elements of market activity, not external constraints on it. The market in this tradition is a space for civic relationships, not only for transactions. This tradition did not disappear; it persists in cooperative structures, artisan networks, and local institutions across Europe.

14. Civil economy organizations integrate exchange, redistribution, and reciprocity simultaneously.
The civil economy approach is not a "softer" version of capitalism. It is a different operational logic that combines: the principle of equivalent exchange (standard market), the principle of redistribution (social equity mechanisms), and the principle of reciprocity (mutual obligation between parties). Businesses that only operate on exchange logic are a subset of what markets can be, not the default definition of them.

15. Replication across locally-owned implementations is more durable than centralized scaling.
A method that enables many independently-owned operators to replicate a proven model is more resilient — and more aligned with civil economy principles — than a single company that captures a large market. Governance, decision-making, and accountability stay local. Network effects accumulate without monopolization. This is the structural logic of cooperatives, franchise networks with shared values, and movements — not of platforms.

Open Questions (Not Conclusions)

These principles are not yet connected into a business. The connections that remain unresolved:

• What specific human has a problem acute enough to pay to solve it tomorrow?
• Which two or three of these principles combine into a single value proposition?
• Who are the co-builders, and what do they bring that is complementary?
• What is the revenue mechanism in the first six months?
• What specific thing in the world gets measurably better because this exists?

This document is designed to be read by anyone without prior context. It is a foundation for further thinking, not a pitch.