Forget the "blue wave" for a moment. The contest between Democrats and Republicans is the least interesting part of what happens after the midterms. The deeper question is whether the United States, with governments that legitimately change every two or four years, can build a long-term vision for the country and compete with the planned economy of China and, to a lesser extent, the wider Global South, the BRICS and the countries tied to the Belt and Road Initiative.
Artificial intelligence and semiconductors are the test case. These are not industries that reward a four-year horizon. Chips depend on supply chains, talent, energy and capital commitments that mature over decades. A strategy that is launched by one administration, rebranded by the next and reversed by the one after is not a strategy. It is a series of slogans with a budget attached.
China does not have this problem. Its Party guides the economy, society and the moral direction of the country with a very long view, and it does not have to look over its shoulder at the next round of local elections. We have seen what that means in renewables: for three years running, China has installed more solar capacity than the rest of the world combined, and in 2025 it accounted for 57% of all new solar installed worldwide, according to SolarPower Europe. Nobody should be surprised. When planning is not interrupted by electoral calendars, a country can look beyond the next vote and organise its industrial independence. The same logic applies to AI. China has been investing in it for years, and its new language models already appear to rival the best American ones: by February 2025, Stanford's AI Index found, the best American model led the best Chinese model by only 1.70% on the Chatbot Arena leaderboard.
America's answer so far has been impressive in size and fragile in structure. A $500 billion, four-year investment plan for AI infrastructure was announced in January 2025. But an announcement is not a commitment that survives the next election. The real question is whether the plan outlives the political mood that produced it, or whether AI and chips become hostages of the quarrel between left and right, each side treating the technology as a weapon against the other rather than as a national asset.
The latest news shows how easily this happens. The CEO of Nvidia, Jensen Huang, is to be questioned in the Senate over alleged links to people close to China. Whatever the outcome, the signal is damaging. Doing business in high technology has become extraordinarily delicate precisely when the country needs its most successful technology companies to be confident, ambitious and globally present. We have seen this tension before: when Gulf states bought Nvidia chips, the headlines worried less about the commercial success than about what "autocratic" governments might do with them. Markets cheered, politicians fretted, and no one explained how a coherent strategy could be built on both reactions at once.
This is the paradox at the heart of the tech war. Washington is waging it against China and, in practice, against much of the rest of the world. But a war fought with export controls, suspicion and public hearings has a cost, and the first bill lands on American companies, American investors and American researchers; Nvidia alone booked a $4.5 billion charge in the quarter ended April 2025 after Washington required export licences for its H20 chips. I fear that the United States is, in the end, also fighting against itself.
This does not mean that democracy is doomed or that the Chinese model is simply superior. It means that democracies must be honest about their structural weakness and design around it. If governments change every few years, then the long-term elements of strategy, such as chips, energy, skills and basic research, need a degree of continuity that does not depend on who wins. That requires a serious debate about where to draw the line between State and Market: which sectors, which geographies, which supply chains, how to compete with China and with the rest of the world, how to handle AI and economies of scale. It requires a holistic approach, and that is exactly what is missing.
The midterms will be analysed endlessly as a verdict on one party or one president. That is the wrong lens. The verdict that matters will be delivered in a decade, in fabs, data centres and laboratories, and it will go to whoever managed to keep a plan alive across many elections. The United States has the talent, the capital and the companies. What it must prove is that it has the patience.
Leave a Reply