Peter Thiel: AI could be bigger than the internet; Europe is 'extremely far' behind

Investor Peter Thiel believes artificial intelligence could surpass the internet and kick off a new era of technological growth. At the same time, the PayPal and Palantir cofounder warns that Europe is, in his words, “extremely far” behind the United States on AI. He says so in an in-depth interview with Mathias Döpfner.

Thiel has long been known for his so-called stagnation thesis. In his view, technological progress since roughly the 1970s has been far less impressive than commonly assumed. Computers, software, and the internet advanced at breakneck speed, but in sectors like energy, space, chemicals, and heavy industry, real progress lagged.

AI has partly changed his mind.

“Perhaps AI is bigger than the internet,” Thiel says during the conversation. He argues artificial intelligence at least has the potential to re-accelerate economic and technological growth. He doesn’t see AI as just the next wave of software, but as a technology that could reshape broad swaths of the economy.

Today’s AI debate often centers on whether the technology is moving too fast. Thiel thinks there’s not enough focus on the opposite scenario: a world where progress is deliberately slowed.

He doesn’t dismiss extreme AI risks entirely. In the interview, he discusses a scenario where a system becomes far more intelligent than humans and potentially slips beyond human control. He says it’s nearly impossible to assign reliable probabilities to that.

Even if the risk were relatively small—say 5 or 10 percent—Thiel argues it’s still serious enough to warrant attention.

But that doesn’t automatically mean development should be slowed.

Thiel contends that prolonged economic stagnation also creates major social risks. Modern societies, he says, are partly built on the expectation that children will be materially better off than their parents. When that progress stalls, economic and political conflict can intensify.

He’s pushing back on the idea that less technological progress is automatically a safer or more stable path.

Thiel sees geopolitics as another sticking point. If one country slows AI on purpose while others charge ahead, the balance of technological power simply shifts, he argues.

China comes up prominently in the discussion. Thiel says a truly coordinated slowdown would only be possible if virtually all major countries moved in lockstep.

That would require an exceptionally powerful form of international cooperation. He even floats the hypothetical of a kind of world government capable of enforcing such rules—something he sees as potentially a bigger problem than the technology it would regulate.

The underlying question is familiar in today’s AI debate: how effective can national rules be when models, chips, data centers, and research are global?

Thiel is particularly critical of Europe.

He describes the United States as the place where the most important AI work is happening right now. China, he says, is a fast follower. Europe, in his view, is much further back.

“Europe is extremely far behind,” Thiel says.

The gap, he argues, isn’t just visible among companies building foundational AI models. In the U.S., traditional enterprises also feel far more urgency to actually integrate AI into their operations.

Thiel points to Palantir, the data company he co-founded. Demand from U.S. firms for Palantir’s technology has surged in recent years, he says, as businesses try to figure out how to adapt to the AI era.

In Europe, he claims, that urgency is far harder to find.

Later in the conversation, Thiel uses two historical examples to explain Europe’s position: Concorde and Minitel.

The Concorde was the supersonic passenger jet jointly developed by the United Kingdom and France. Although Europe initially trailed the United States in supersonic aviation, Thiel notes it still managed to deliver an ambitious technological project.

For him, Concorde shows how Europe can turn a technological deficit into a breakthrough.

France’s Minitel represents the opposite.

Minitel was a French digital network that offered electronic services to households before the World Wide Web. France tried to build its own digital ecosystem, but the closed system was ultimately overtaken by the open internet.

Thiel then poses the question: which of the two scenarios best fits European AI?

His verdict is blunt: in his view, Europe is currently even further behind in AI than France once was in the early days of the internet.

Thiel sees part of the answer in Europe’s entrepreneurial culture.

In Germany, he notes, people often talk about a fear of failure. He suspects there may also be a “fear of success.”

European founders, he argues, are less likely to scale successful companies into massive global players. Once a business is doing well, it’s more likely to be sold than aggressively scaled up.

Thiel backs this with a striking comparison between the richest Americans and Germans.

Among younger ultra-wealthy Americans, a large share built their fortunes by founding companies. Among younger Germans in the same wealth bracket, the figures he cites suggest fortunes are overwhelmingly inherited.

For Thiel, that symbolizes a bigger problem: Europe isn’t creating enough new enterprises that can grow into world-class companies.

The debate is about more than ChatGPT or large language models.

AI, Thiel argues, exposes a longstanding issue: the United States has an ecosystem where new tech firms can scale to enormous size quickly, while Europe has produced far fewer companies that dominate the world’s digital infrastructure.

That doesn’t mean Europe has already lost the AI race. The landscape is still shifting, and European companies, governments, and research institutions are stepping up investments in homegrown AI models, chips, supercomputers, and data centers.

But the sheer scale of U.S. investment makes the gap ever more visible.

Companies like Microsoft, Google, Amazon, and Meta are pouring tens of billions into AI infrastructure. Many of the most important models and AI platforms also come from U.S. players such as OpenAI, Anthropic, Google, and Meta.

Europe does have contenders like Mistral AI, but it still lacks an ecosystem of comparable scale.

Thiel’s Concorde and Minitel analogy neatly captures Europe’s challenge.

Europe can try to build its own AI industry and close the gap. But if investment, compute, and scale lag, there’s a real risk European firms will end up primarily as users of technology developed elsewhere.

For Thiel, AI isn’t just another technological revolution. It’s a test of whether Europe can still build large-scale technology industries of its own.