第一财经

Are tech leaders hesitating? Trump doesn't think so: The US needs to take the lead

原文:科技领袖们在犹豫?特朗普不以为然:美国要领先

Trump's View on the "AI Survival Crisis": A Threat to Existence and a Lifeline for the Economy

Hello everyone, I'm your financial journalist. Recently, the American political and tech circles have been in a heated debate, with one central question: Should we put the brakes on artificial intelligence (AI)?

On one side, tech giants like OpenAI and Anthropic are freaking out, claiming that the rapid development of AI could lead to human extinction, and they are urging the government to take action, even suggesting the need for a "third-party arbiter" to oversee the industry.

On the other side, the Republican government led by Trump believes: "Don't panic, don't stop us; we need to move fast, and the fastest will be the leader."

This is not just a debate about technological approaches; it's also a fierce battle over national security, economic interests, and regulatory power. Today, we'll break down this news in simple terms to uncover the underlying complexities.

---

1. Trump's "Guardrails Theory": Don't Get in the Way, I Want to Win

First, let's clarify Trump's current stance. In an interview in Ireland, he said clearly: **"Whoever wins AI wins the future."

Trump is somewhat dismissive of the safety warnings and regulatory demands from tech giants. His logic is simple:

  • Guardrails are acceptable, but not walls: You can set up basic rules to prevent cars from falling off a cliff, but you can't ban cars from speeding just because you're afraid of accidents.
  • Maintaining leadership is essential: The U.S. must stay ahead in AI. If excessive regulation allows China or other competitors to overtake us, that would be the real danger.
  • Doubting the severity of the threat: Trump even suggested that those who warn of human extinction might be exaggerating; such extreme scenarios are "impossible."

In plain language: It's like two racers. Driver A (the tech giants) says, "This car is too fast and might crash; let's put a speed limit." Driver B (Trump/the government) replies, "No way, the finish line is ahead; if we put a speed limit, we'll lose. You can check the brakes, but don't make me slow down."

Republican lawmakers behind Trump (such as House Speaker Johnson) agree, arguing that tech companies have the ability to control their own pace and don't need government interference. Some even mock that if tech companies negotiate regulation in exchange for not developing super-intelligent systems, it's like "extorting the public."

---

2. Tech Giants' "Panic" and "Self-Saving Measures": We're Really Scared

Why have the tech giants suddenly called for a halt? It's not just for show; they're genuinely concerned:

  • Insider Revelations: After Anthropic researcher Coxon resigned, she publicly warned that AI developers believe this technology could lead to human extinction by the end of the century.
  • CEO Appeals: Anthropic's CEO Amodei, OpenAI's CEO Altman, and even Elon Musk have rarely agreed, calling for a slowdown.
  • Specific Demands: They want to establish industry safety standards, introduce third-party evaluations, and unify global regulations.

In plain language: It's like a group of chefs developing a new, potent dish. They realize it might be dangerous, so they say to customers and regulators, "This dish is too strong; we need to test its safety and set standards before mass production."

However, this approach doesn't work on the Trump administration. They think, "If you're so capable, why can't you control it yourself? Why do you need the government to tell you how to cook?" This conflict reflects the clash between inherent technological risks and external regulatory authority.

---

3. Europeans' "Envy" and "Anxiety": Why Do They Flock to the U.S.?

The news mentions an interesting phenomenon: Many European business school students want to start companies or invest in the U.S. INSEAD professor Michoff pointed out, "As a founder, it's much easier to make money in the U.S. than in Europe."

Why?

  • Different Regulatory Environments: Europe has strict regulations (e.g., GDPR) and high startup barriers, while the U.S. is more lenient and encourages innovation.
  • Big Gap in Computing Power: Former Italian Prime Minister and ECB President Draghi warned that EU AI power accounts for less than 5% of the global total, compared to 75% in the U.S. This means Europe is far behind in AI infrastructure.

In plain language: It's like two countries competing in a race. The U.S. offers the best tracks and fastest fuel stations (computing resources) with flexible rules, allowing the best drivers (talents) to excel. Europe has narrower tracks, stricter rules, and fewer fuel stations, so top talents and capital flow to the U.S. Europeans are anxious but can't change this in the short term.

---

4. Is AI a "Lifesaver" or a "Bubble"? The Economic Calculations Aren't Clear

This is the core issue. Why does the Trump administration support AI so much? Because the U.S. economy is facing challenges:

  • High Inflation + Weak Growth: Inflation hasn't subsided, and economic growth is sluggish.
  • **AI as a "Cure": Fed Chair Powell and others believe AI will boost productivity, reduce inflation, and increase wages.
  • Revelations from the Oxford Economics Institute: AI only boosts U.S. consumption by 0.4% annually and 0.1% in other countries. Investment growth (3% annually) is largely driven by AI, but without it, other sectors are shrinking (-2%).
  • Productivity No Boost: Despite claims of increased productivity, adjusted productivity indicators show no improvement.

In plain language: It's like a company with struggling traditional businesses but a profitable AI division. The boss (the government) claims, "Our company is doing well because of AI!" But the real situation shows declining traditional businesses and an AI division that's costly and hasn't significantly improved efficiency.

Worse, AI companies are heavily in debt. Reports predict that AI-related bond issuance could reach $500 billion by 2026. If AI doesn't generate real revenue, these debts could become a financial time bomb.

---

5. Who Will Be the Winners in the Future? From "Selling Shovels" to "Mining Gold"

Let's look at what investors and experts think about the future:

  • Current Stage: Winners are those selling the necessary infrastructure (chips, computing power, data centers).
  • Future Stage: Winners will be those who convert AI into actual revenue and growth.

MFS Investment Management's Anna notes that investors are now distinguishing between these two types of companies. They used to buy AI stocks blindly; now they're more selective: "How much money will your AI really generate for me? Or is it just burning money?"

In plain language: It's like a gold rush:

  • Now: Everyone buys shovels, jeans, and buckets (chips, servers) because demand is high, but if no gold is found, the shovels will be unsold.
  • Future: The market will favor companies that actually create value through AI applications. If AI doesn't save costs or generate revenue, the current infrastructure bubble could burst.

---

Summary

This news reveals a complex situation:

1. Politically: The Trump administration aims to maintain U.S. technological dominance with minimal regulation and skepticism about safety concerns.

2. Technologically: Tech giants seek self-regulation and external oversight but are seen as obstacles to innovation.

3. Economically: AI is a cover for the U.S. economy's weaknesses, masking declining traditional industries. While it drives investment and consumption, it hasn't significantly boosted productivity and relies on debt.

4. Globally: The U.S. is using lax regulations and strong computing power to dominate Europe, shifting the global AI landscape in its favor.

For the public: Don't blindly believe in AI's omnipotence. While its potential is vast, it's mainly a capital game and an infrastructure race. Investors should be cautious of companies that only burn money. For individuals, it's more important to see how AI will change jobs and incomes. Ultimately, only those who turn AI into real productivity will be the winners.