A Rare Confession from the Epicenter of the Revolution
On March 11, 2026, at BlackRock's Infrastructure Summit — the world's largest investment fund with approximately 11.5 trillion USD in assets under management — OpenAI CEO Sam Altman said something technology leaders typically avoid stating outright: AI (Artificial Intelligence) is breaking the balance of power between labor and capital, and nobody knows what to do about it.
This is not a warning from a critic or pessimistic academic. This is a confession from the head of the company leading the AI revolution — someone with direct financial interest in convincing the world that everything will be fine. When Altman himself says "it frankly stumps him," that is a signal that everyone — from programmers in Silicon Valley to nail salon owners in Little Saigon — needs to stop and listen.
For the Vietnamese American community in the United States, a community whose livelihoods are tied to both manual labor and knowledge services, what Altman is saying is not science fiction. It is happening right now.
From "Scarcity" to "Surplus": A Problem Capitalism Has Never Had to Solve
Altman cites an idea that he says has haunted him: humanity spent thousands of years building social structures to manage scarcity, and now must rapidly learn to manage surplus. It sounds like a luxurious paradox — who complains about having too much? But the economic logic behind it is deadly serious.
Capitalism operates on a fundamental premise: labor has value because it is scarce. You work, you get paid, you use that pay to buy goods and services. This cycle — labor creates income, income creates consumption, consumption creates jobs — is the backbone of every modern economy. But if a GPU (graphics processing unit, now the primary computational unit for AI) can do the work of a knowledge worker better at a cost hundreds of times lower, that cycle breaks.
What is noteworthy is that Altman is not talking about a vague future. He states that AI has already crossed the threshold of "major economic utility" in recent months, evolving from code-writing assistance to performing complex tasks across many knowledge fields. He predicts that very soon, AI agents will be trusted to handle jobs that span many days, many weeks — autonomous operations like a senior employee.
Real-world data supports this assessment. According to the McKinsey Global Institute report updated in January 2026, approximately 30% of work hours in the United States could be automated using current AI technologies — a significant increase from the 21% estimate in 2023. The hardest-hit sector is not manual labor but knowledge work: data analysis, legal drafting, preliminary design, translation, and basic programming.
"AI Washing" and the Blurred Line Between Real and Fake Fear
One of the most subtle points in Altman's remarks is the distinction between inflated AI fears and legitimate AI fears — and the acknowledgment that both are happening simultaneously.
The phenomenon of "AI washing" — companies blaming AI for layoffs when the real cause is poor business, restructuring, or simply cost-cutting — has become so common that President Donald Trump had to warn that AI is facing a "serious public relations problem." When every round of layoffs comes with a press release mentioning AI, the public naturally reacts with confusion and anger.
But here lies the fundamental dilemma: even if 50% of recent layoffs dishonestly invoke AI as an excuse, the other 50% might be real. And the long-term trend cannot be denied. Altman points out that a new generation of startups is deliberately avoiding hiring many people, instead channeling capital into computational power. In India, he observes entrepreneurs building "zero person startups" — relying entirely on AI for software writing, legal processing, and customer support management.
If this model succeeds and spreads, the implications are profound: future economic growth could become entirely decoupled from job growth. GDP rises, profits rise, but the number of well-paying jobs falls. This is not a fictional scenario — it is already emerging in the technology sector.
The Vietnamese American Perspective: Who Gets Hit First?
For approximately 2.3 million Vietnamese Americans, this story touches multiple livelihood layers.
Layer 1: Knowledge work and technology. The Vietnamese American community has a higher participation rate in STEM fields than the national average. According to Census data, approximately 33% of second-generation Vietnamese American workers are employed in highly specialized fields, including software engineering, data analysis, and accounting. These are precisely the fields where Altman says AI is advancing most rapidly. Young Vietnamese American engineers in San Jose, Austin, or Seattle are on the front lines: they are both AI users and potentially AI-replaceable.
In particular, junior-level programmers — historically a gateway into the profession for many young Vietnamese Americans fresh out of school — are being squeezed hard. When AI can write code at a junior developer's level, companies have no reason to hire 20 juniors when 5 seniors plus AI will do.
Layer 2: Small business and services. The nail industry — where Vietnamese Americans account for approximately 50% of the US market share with an estimated over 80,000 salons — appears safe from direct automation in the near term, since it requires physical interaction. But indirect impacts are not negligible: if customers lose jobs or see income decline due to AI, they cut luxury service spending first. A contraction in knowledge-work employment will cascade into contraction in personal services.
Moreover, AI is already beginning to change how small businesses operate: automatic scheduling, bookkeeping management, social media marketing. Nail salon and pho restaurant owners who adopt quickly will gain advantage; those who lag will fall behind.
Layer 3: Remittances and trans-Pacific connections. Remittances from the United States to Vietnam reached approximately 14 billion USD in 2025, with a significant portion coming from the Vietnamese American community. If AI triggers the "painful adjustment period" Altman warns of, this remittance stream will be affected — directly impacting millions of families in Vietnam who depend on this money for education, healthcare, and small investment.
Simultaneously, Vietnam's software outsourcing sector — which provides programming services to US companies — faces serious risk if AI replaces junior and mid-level programmers. Many Vietnamese Americans play bridging roles in these outsourcing contracts; if the market shrinks, the trans-Pacific business network shrinks with it.
H-1B Visas and the Immigration Paradox in the AI Age
One understated aspect: if AI truly reduces demand for knowledge workers, political pressure to lower H-1B visa quotas will intensify. The argument will be: why import foreign programmers when AI can replace most of the work they do?
This directly affects the labor-skilled migration flow from Vietnam to the United States — an important channel alongside family-based immigration. According to USCIS data, Vietnam ranks in the top 10 countries whose citizens receive H-1B visas. If this channel narrows, it will reshape the Vietnamese-American immigration picture for the coming decade.
Yet there is an interesting paradox: the very best engineers — those who know how to use AI to multiply productivity — will be more sought after than ever. AI does not only replace; it also amplifies. The gap between the "top 10%" and everyone else will widen, creating a two-speed labor market: superstars earning very high pay, and everyone else struggling.
"Intelligence as Cheap as Tap Water" — Dream or Nightmare?
Altman describes OpenAI's vision as making AI "too cheap to meter" — selling artificial intelligence as a basic utility, like tap water or electricity. The phrase "too cheap to meter" has a memorable history: Lewis Strauss, chairman of the US Atomic Energy Commission, used it in 1954 to describe the future of nuclear power. Seventy-two years later, nuclear power has never become that cheap.
But even if Altman succeeds, the nagging question remains: in a world where artificial intelligence is as cheap as tap water, how do people make money? If you can hire an AI agent for a few dollars per hour to do work that previously cost 50 dollars per hour to employ a human, where does that economic surplus go? Into the business owner's pocket. Into shareholders' accounts. Into the hands of those who own GPUs.
This is the essence of the "labor-capital imbalance" Altman acknowledges. Historically, labor had negotiating power because nothing could replace it. Unions, labor law, minimum wage — all rest on the premise that businesses need workers. When that premise weakens, the entire legal and social framework built on it wavers.
Altman mentions that GDP could "plummet" in a "permanent deflationary world." This is the scenario the shocking report from Citrini Research warned about: when AI reduces production costs close to zero, prices fall, revenue falls, GDP measured in money falls — even though actual output rises. The economy balloons in physical terms but contracts financially. Altman calls this "ghost GDP." This is not conspiracy theory; it is basic economics when you push the marginal cost of knowledge production toward zero.
Altman Is Both Doctor and Disease Vector
The greatest irony in this entire story: Altman stands on the BlackRock stage — a podium representing the apex of capital — to confess that his product is tilting the balance toward capital and against labor. Simultaneously, he calls for building colossal infrastructure (gigawatt data center campuses) to accelerate this process. He partners with North American construction unions to build data centers — using manual labor to construct the infrastructure that will replace knowledge work.
Altman positions himself as both doctor and disease. He diagnoses precisely the problem he is creating, but the only solution he offers is... to keep creating the problem, in hopes that eventual surplus will benefit everyone. That is Silicon Valley's "move fast and break things" logic, but this time what breaks might be the economic structure of society itself.
Demis Hassabis of Google DeepMind, Altman's rival, is more optimistic, speaking of a "new Renaissance" — but even he admits there will be a "shakeout" lasting 10 years. For a family raising children, paying mortgages, and sending money back to Vietnam, 10 years is not a "transition period." It is an entire generation.
Conclusion: Preparing for Something Nobody Knows How to Prepare For
Sam Altman is right about one thing: nobody has the answer ready. But that confession, coming from the person creating the problem, should be treated as a warning bell rather than reassurance.
For the Vietnamese American community, the practical lesson is clear: skill diversification is the best insurance. Those working in technology need to move toward the "using AI" side rather than the "being replaced by AI" side. Those running small businesses need to start exploring AI as a tool, not a distant threat. And at the community level, Vietnamese American organizations in the US need to start serious conversations about economic futures — not just for this generation, but for the generation growing up in a world where artificial intelligence is as cheap as tap water.
Ultimately, the biggest question is not what AI can do. The question is: when AI can do nearly everything, where does human value lie? Altman says he does not know. But history shows that those who prepare earliest — even if they prepare in the wrong direction — still have advantage over those who do not prepare at all.
The game has changed. The fact that nobody knows the new rules does not mean you get to sit out the game.