Closing: The Chessboard

Ten episodes on the geopolitics that will define European industry for a generation — and a note from the author

By VastBlue Editorial · 2026-03-26 · 15 min read

Series: The Chessboard · Episode 11

The board, at a glance

Over ten episodes, this series has mapped the geopolitical forces that will determine which industries exist in Europe in 2035. Not which industries Europe would like to have. Not which industries its strategy documents envision. Which industries the interaction of American capital, Chinese manufacturing scale, and European regulatory architecture will permit to survive and grow on European soil. The distinction is not rhetorical. It is the difference between a wish and a forecast.

The Chessboard began with a premise: that geopolitics is not background noise for European industry. It is the operating environment. Every company building in Europe today — whether it manufactures batteries, deploys AI systems, generates electricity, or designs semiconductors — operates inside a set of constraints that were not created by market forces alone. They were created by political decisions made in Washington, Beijing, and Brussels, by trade agreements and trade wars, by subsidy programmes and export controls, by energy policies and defence budgets. Understanding those constraints is not optional. Misunderstanding them is expensive. Ignoring them is fatal.

What follows is not a summary. Summaries flatten. This is an attempt to draw the through-lines — to show how ten apparently separate episodes connect into a single argument about where Europe stands, what it risks, and what it must do.

The three industrial strategies

Episode 1 — "The new map of industrial power" — established the frame. Three blocs, three strategies, one future. The United States chose to spend: $369 billion in IRA subsidies, uncapped tax credits designed to pull manufacturing investment toward American geography through the gravitational force of direct cost reduction. China chose to build: two decades of state-directed capital, coordinated across provincial governments, state-owned banks, and national champions, producing manufacturing overcapacity so vast that it functions as a strategic weapon. Solar panels at $0.15 per watt. Electric vehicles at $10,000. Battery cells at prices the global industry did not expect to see until 2030. Europe chose to regulate: CBAM, the Green Deal Industrial Plan, the AI Act, the Digital Markets Act — instruments of normative power deployed by a bloc whose institutional architecture makes direct industrial subsidy structurally difficult.

The argument of that first episode has been vindicated by every subsequent one: in a contest between money, coordination, and rules, rules finish third unless they are backed by the industrial capacity to make them matter. A carbon border adjustment is meaningful only if Europe manufactures the clean-energy products it needs. An AI regulation is meaningful only if Europe hosts the AI infrastructure it governs. A defence procurement preference is meaningful only if Europe produces the weapons systems it requires. Standards without production capacity are suggestions, not strategy.

What made the first episode's framing durable was its refusal to treat these strategies as morally equivalent choices in a policy seminar. They are not. They are competitive instruments deployed by blocs with different structural advantages and different theories of how power works. The United States has deep capital markets and a reserve currency that allows it to borrow at low cost — so it spends. China has state-directed capital and a political system that sustains strategic patience across decades — so it builds. Europe has regulatory sophistication and institutional credibility — so it regulates. Each strategy leverages what the bloc actually has. The question is whether what Europe has is sufficient for the contest it is in.

Dependencies that cannot be wished away

Episodes 2 and 4 examined the physical and energetic substrates of industrial competition — the dependencies that no amount of strategic ambition can override without addressing directly. "Chips, batteries, and the things you cannot buy" traced Europe's critical supply chain vulnerabilities: near-total dependence on Taiwan for advanced semiconductor fabrication, overwhelming reliance on China for battery cell production, and a rare earth processing chain that runs almost entirely through Chinese refineries. These are not market inconveniences. They are strategic vulnerabilities of the first order. When TSMC's fabrication lines in Hsinchu represent over 90 percent of the world's most advanced chip production, and when those fabrication lines sit 130 kilometres from a military power that considers the island a renegade province, the word "dependency" understates the reality. The word is "exposure."

"The energy equation nobody wants to solve honestly" confronted the fact that European industrial electricity prices are two to three times higher than American ones and four to five times higher than Chinese ones. This is not a minor cost differential. It is a structural disadvantage that compounds across every energy-intensive manufacturing process — steel, aluminium, chemicals, cement, glass, semiconductor fabrication, data centre operations. A European aluminium smelter paying €80 per megawatt-hour competes against a Chinese smelter paying €15 and an American one paying €30. The physics does not change because the policy ambitions are admirable. Electrons do not care about strategy documents. Factories follow electrons, and Europe's electrons are expensive.

Factories follow electrons, and Europe's electrons are expensive. No industrial strategy survives contact with an electricity bill that is three times your competitor's.

Episode 4 — The energy equation nobody wants to solve honestly

The nuclear question — which Episode 4 confronted directly — encapsulates Europe's difficulty with energy-intensive industrial decisions. France, which maintained its nuclear fleet, has among the lowest industrial electricity prices in Western Europe and a largely decarbonised grid. Germany, which closed its last nuclear plants in April 2023 while simultaneously trying to exit coal, faces structurally higher electricity prices and a grid that still depends on natural gas — gas that, until February 2022, came primarily from Russia through pipelines that are now damaged or shut. The divergence is not an accident. It is the cumulative consequence of political decisions made over decades, and it cannot be reversed in years. Energy infrastructure operates on timescales measured in decades. A nuclear plant takes ten to fifteen years from decision to first electricity. A grid transformation takes a generation. The energy disadvantage Europe faces today was determined by decisions made — and decisions avoided — between 2000 and 2015.

The instruments of control

Episodes 3, 5, and 6 examined the mechanisms through which geopolitical competition is conducted: AI systems, trade instruments, and military procurement. Each revealed a different facet of the same structural challenge.

"AI sovereignty is not about building another ChatGPT" made the argument that Europe's AI debate has been systematically misdirected. The public conversation fixates on foundation models — on whether Europe can build a competitor to GPT-4 or Gemini. This is the wrong question. The strategically consequential AI is not the chatbot. It is the system managing electricity grid balancing, the model optimising manufacturing quality control, the algorithm routing military logistics, the platform processing hospital diagnostic imaging. Industrial AI — AI embedded in the physical systems that society depends on — is where sovereignty actually matters, because it is where loss of control has consequences measured not in market share but in grid blackouts, production failures, and compromised defence systems.

The distinction is critical because it reframes what European AI sovereignty requires. Europe does not need to compete with OpenAI on foundation models — a race it would likely lose and does not need to win. It needs to ensure that the AI systems operating inside its critical infrastructure are developed, deployed, and maintained by entities subject to European oversight. This is an achievable objective. It requires investment in applied AI engineering, in domain-specific training data, in secure compute infrastructure, and in the institutional frameworks that govern how AI is used in sensitive systems. It does not require building a European GPT. It requires building European competence in the AI that actually runs things.

"When trade policy becomes industrial policy" catalogued the proliferating instruments of economic coercion: carbon border adjustments, semiconductor export controls, forced localisation requirements, procurement restrictions, investment screening mechanisms, and sanctions regimes. The October 7, 2022 semiconductor export controls — which restricted the export of advanced chip-making equipment and technology to China — represented perhaps the most consequential use of trade restrictions as a tool of industrial destruction since the Cold War. The controls were not designed to slow Chinese semiconductor development. They were designed to stop it. And they revealed a principle that every subsequent episode reinforced: trade policy is no longer about comparative advantage and mutual benefit. It is about strategic denial and industrial positioning. The rules-based trading order that Europe championed and benefited from for decades is being replaced by a system in which trade instruments serve security objectives, and security objectives are defined expansively enough to encompass entire technology sectors.

"Europe's defence awakening and the industrial base it needs" made the connection between security spending and industrial capacity explicit. NATO's 2 percent GDP defence spending target — once treated as aspirational — has become a floor, with serious discussion of 3 to 3.5 percent targets. But spending targets are meaningless without production capacity. Europe cannot buy the ammunition, the armoured vehicles, the naval vessels, the air defence systems, and the military satellites it needs if European factories do not produce them. And European defence-industrial capacity has been eroded by three decades of post-Cold War disinvestment. Rheinmetall cannot triple its ammunition output overnight. Naval Group cannot build frigates faster than its shipyards and supply chains allow. Airbus Defence cannot scale helicopter production without expanding a supplier network that stretches across dozens of countries. Defence spending is industrial policy, whether it is acknowledged as such or not — and the industrial base it requires overlaps substantially with the civilian industrial base that every other episode in this series has examined.

The human and institutional dimension

Episodes 7, 8, and 9 turned from hardware and policy to the human and institutional factors that will determine whether Europe can execute any strategy at all.

"The talent war Europe is losing quietly" documented the brain drain that few European policymakers discuss with the specificity it demands. Europe produces approximately 30 percent more STEM graduates per capita than the United States. It retains a significantly smaller proportion of them in high-value technology roles. The salary differential is part of the explanation — a senior AI engineer at Google in Mountain View earns two to three times the compensation of an equivalent engineer at a European institution. But salary is not the whole story. The deeper issue is the density of consequential work. Silicon Valley, for all its mythologies, offers engineers the opportunity to work on systems that operate at global scale, with compute budgets that European institutions cannot match and data access that European regulations actively restrict. The talent leaves not only because the money is better but because the problems are bigger — or at least, the resources available to solve them are.

The irony is structural. Europe invests heavily in producing world-class engineers — through publicly funded universities, research institutions, and training programmes that American students would envy — and then watches a disproportionate share of that investment walk through airport departures at Schiphol, Frankfurt, and Heathrow. The return on public investment in STEM education accrues to American technology companies, who recruit the graduates, and to American shareholders, who capture the value they create. This is not a market failure in the conventional sense. It is a system operating exactly as designed — a system in which talent flows freely to where the compensation, resources, and opportunities are greatest. The failure is in Europe's inability to make itself that destination for the engineers it trains.

"Small countries, disproportionate strategies" offered a counterpoint — and, implicitly, a model. Estonia, Israel, Switzerland, Portugal, Taiwan: five nations that have built outsized influence through precision rather than scale. Estonia digitalised its entire government. Israel built a defence-technology ecosystem that exports globally. Switzerland positioned itself at the intersection of pharmaceutical innovation, precision manufacturing, and financial infrastructure. Portugal leveraged its geographic and cultural position to attract digital talent and forge connections across the Lusophone world. Taiwan made itself indispensable by concentrating on a single technology — semiconductor fabrication — and executing with a discipline that larger nations could not match. The common thread is strategic clarity: knowing what you are, knowing what you are not, and concentrating resources on the intersection of national capability and global demand.

"The EU's next five years will define its next fifty" synthesised the institutional challenge. The 2024-2029 European Commission faces a policy agenda of unprecedented density: industrial policy, digital regulation, AI governance, defence integration, energy transition, enlargement, and institutional reform — simultaneously. The Draghi Report called for €750 to €800 billion per year in additional investment. The Letta Report called for completion of the single market in capital, services, energy, and digital. The Franco-German expert group called for institutional reforms — including expanded qualified majority voting — necessary to accommodate an EU of 30-plus member states. Each report was credible. Each was ambitious. And each demanded a level of political consensus and institutional capacity that the EU has historically struggled to achieve even on single-issue reforms, let alone across multiple fronts simultaneously.

The single argument

Taken together, these ten episodes make a single argument. It is this: Europe possesses every ingredient required for industrial competitiveness in the 21st century — scientific talent, engineering depth, a wealthy consumer market, institutional credibility, regulatory sophistication, and a quality of life that should, in theory, attract and retain the best minds in the world. What it lacks is not capability but velocity. The speed at which strategic decisions are made, funded, and executed. The speed at which permits are granted, factories built, and supply chains assembled. The speed at which capital moves from risk-averse institutional allocators to the companies and projects that need it.

The velocity deficit shows up in every domain this series examined. In semiconductor fabrication, TSMC announced and began constructing its Arizona plant in 2020; the European Chips Act was proposed in 2022 and its signature projects remain in planning or early construction. In battery manufacturing, CATL and BYD have built dozens of gigafactories across China and are expanding into Hungary, Morocco, and Brazil; Northvolt, Europe's most prominent battery startup, has struggled with production yields at its flagship Swedish facility. In AI infrastructure, Microsoft, Google, and Amazon have each committed over $50 billion in capital expenditure on data centres and compute capacity in a single year; Europe's collective AI infrastructure investment remains a fraction of any one of those figures. The pattern is consistent: Europe identifies the strategic imperative correctly, announces the policy response competently, and executes more slowly than the competition.

This is not a new observation. Mario Draghi said it. Enrico Letta said it. Emmanuel Macron has said it. Every serious analyst of European competitiveness arrives at the same diagnosis. The question is whether diagnosis leads to treatment — and whether treatment arrives in time. Because the geopolitical environment mapped in this series is not static. It is accelerating. The United States is spending faster. China is building faster. Export controls are tightening. Supply chains are fragmenting. Alliance structures are shifting. The window for European action is not closing gradually. It is closing on a schedule set by decisions made in Washington and Beijing, over which Brussels has limited influence.

Europe possesses every ingredient required for 21st-century industrial competitiveness. What it lacks is not capability but velocity — the speed at which decisions are made, funded, and executed.

Series thesis

The chessboard metaphor that gives this series its name is imperfect in one important respect. In chess, players alternate moves. In geopolitics, they move simultaneously — and the pace of play is not equal. A player who deliberates while opponents act is not being thoughtful. That player is falling behind. Europe's institutional architecture — consensus-driven, multi-layered, respectful of subsidiarity, protective of national prerogatives — produces deliberation of exceptional quality. It does not produce speed. And in the geopolitical environment of the mid-2020s, speed is not a luxury. It is a survival trait.

None of this means that Europe is doomed. Determinism is lazy analysis. Europe has structural advantages that neither the United States nor China can replicate: a regulatory framework that global companies adapt to rather than ignore, a social contract that produces political stability and quality of life that attracts talent (when immigration systems cooperate), a research base that generates foundational science at world-class levels, and a manufacturing tradition — in automotive, aerospace, chemicals, precision engineering, medical technology — that represents centuries of accumulated competence. These are not minor assets. They are the raw material of industrial power. But raw material is not finished product. The question is whether Europe can convert its assets into industrial outcomes at the speed the moment demands.

A note from the author

We are VastBlue Innovations. We build agentic AI systems for core industries — energy, utilities, manufacturing, financial services — from Funchal, Madeira.

We wrote The Chessboard because the geopolitical forces examined in this series are not abstractions to us. They are the operating environment in which we build. When we deploy AI systems for European energy companies managing grid infrastructure, the electricity pricing dynamics of Episode 4 determine our clients' economics. When we build data integration platforms for European manufacturers, the supply chain vulnerabilities of Episode 2 shape what our clients need. When we hire engineers, the talent dynamics of Episode 7 determine who we can recruit and retain. The geopolitics is not background. It is foreground — as real and as immediate as a line of code or a quarterly revenue target.

The editorial methodology behind this series is the same one that drives our engineering practice. We read the primary sources — the actual text of the Inflation Reduction Act, the Bureau of Industry and Security export control regulations, the Draghi Report in full, the EU defence procurement data, the IEA energy statistics, the OECD migration databases. We do not summarise summaries. We do not cite analyses of analyses. We go to the source, because the source contains information that each layer of intermediation systematically loses. This is true whether you are analysing a geopolitical trend or debugging a data pipeline. The primary source is always more instructive than the commentary.

This was Series 3 — The Chessboard. Ten episodes on the geopolitics that will shape European industry for a generation. If it has changed how you think about one trade policy, one supply chain, one energy statistic, or one institutional constraint, it has done its work. If it has made you impatient with analysis that treats geopolitics as background noise rather than operating environment, it has done more than its work.

What comes next: How We Got Here

Series 4 — How We Got Here — begins next month. Where The Chessboard mapped the present, How We Got Here traces the past. Every constraint examined in this series — every supply chain dependency, every energy pricing structure, every institutional limitation, every talent flow pattern — is the product of decisions made years or decades ago. Some were deliberate. Some were accidental. Some were the result of choosing not to choose, which is itself a choice with consequences.

How We Got Here will examine the historical turning points that created the world this series has mapped. The Marshall Plan and the reconstruction choices that shaped Europe's post-war industrial architecture. The oil shocks of the 1970s and the energy policy divergences they triggered — divergences whose consequences, as Episode 4 of this series demonstrated, are still being paid in industrial electricity bills today. The single-market project and the decisions about what it would — and would not — integrate. The post-Cold War peace dividend and the defence-industrial disinvestment it enabled, which Episode 6 showed now constrains Europe's ability to rearm. The rise of China as a manufacturing superpower and the Western policy responses that ranged from engagement to denial to belated alarm. The creation of the euro and its consequences for industrial competitiveness across a eurozone whose members were never as economically convergent as the currency's architects assumed.

The thesis of the next series is simple: you cannot understand where you are if you do not understand how you got here. The constraints are not natural. They were built — by people, in institutions, through decisions that seemed reasonable at the time and whose consequences were not always foreseen. Understanding those decisions, and understanding why they seemed reasonable, is the prerequisite for making better ones. History is not nostalgia. It is the operating manual for the present.

You cannot understand where you are if you do not understand how you got here. The constraints are not natural. They were built — by people, through decisions that seemed reasonable at the time.

Series 4 preview

Sources

1. Draghi Report — The Future of European Competitiveness — https://commission.europa.eu/topics/strengthening-european-competitiveness/draghi-report_en
2. Letta Report — Much More Than a Market — https://www.consilium.europa.eu/media/ny3j24sm/much-more-than-a-market-report-by-enrico-letta.pdf
3. US Bureau of Industry and Security — Export Control Regulations (October 2022) — https://www.bis.doc.gov/index.php/documents/about-bis/newsroom/press-releases/3158-2022-10-07-bis-press-release-advanced-computing-and-semiconductor-manufacturing-controls-final/file
4. Clean Investment Monitor — MIT and Rhodium Group — https://www.cleaninvestmentmonitor.org/
5. International Energy Agency — World Energy Outlook 2024 — https://www.iea.org/reports/world-energy-outlook-2024
6. European Defence Agency — Defence Data 2023-2024 — https://eda.europa.eu/publications-and-data/defence-data
7. OECD — International Migration Outlook 2024 — https://www.oecd.org/en/publications/international-migration-outlook-2024_b0444c78-en.html
8. European Commission — Carbon Border Adjustment Mechanism (CBAM) — https://taxation-customs.ec.europa.eu/carbon-border-adjustment-mechanism_en