The Wealth Paradox of the AI Era: Why Elon Musk and Jensen Huang's Future Predictions Are Worlds Apart

At the Saudi Investment Forum in November, two tech giants gave vastly different judgments on the job prospects in the AI era. This dialogue seems to concern the future of careers, but essentially reflects the issue of wealth distribution in technological transformation.

Two Contradictory AI Narratives

Elon Musk’s conclusion is quite optimistic: in the next 10 to 20 years, AI and robots will eliminate poverty, everyone will have sufficient material needs, and work will become optional—do it if you want, not do it if you don’t. He uses the analogy of home gardening; although buying ready-made vegetables is cheaper, many people still grow their own out of interest, and the future work model will be similar.

In contrast, Jensen Huang(, CEO of NVIDIA), provides empirical rebuttal: radiologists have not lost their jobs due to AI; instead, hiring demand has increased. The reason is that AI improves diagnostic efficiency, allowing doctors to examine more patients and handle more cases, making them busier rather than idle. This phenomenon reveals a core issue—efficiency improvements do not lead to leisure but instead generate more work.

Warnings from Real Data

Statistics from the U.S. Department of Labor do not support the optimistic outlook. In January, unemployment in the tech industry jumped from 98,000 in December last year to 152,000. According to a survey by ResumeBuilder of 750 business owners applying AI, 37% admitted that in 2023, technology directly replaced jobs, and 44% said layoffs occurred in 2024 due to AI efficiency gains.

The work patterns of lawyers, programmers, and designers already demonstrate this trend. Lawyers using AI to review documents saw their case intake double; developers assisted by AI in programming, and bosses immediately increased feature requirements; designers using generative AI completed what used to take a week in ten minutes, but faced clients demanding twenty versions in a week. Efficiency metrics rise, but human labor becomes even more intensive.

The Flow of Technological Wealth

The fundamental assumption of Musk’s logic is that technological progress will naturally lead to an equal distribution of wealth. However, history does not support this theory. During the Industrial Revolution, people predicted machines would free humans from toil, but workers still worked an average of 16 hours a day; the information age promised paperless offices and shorter hours, yet email made global employees on call 24/7. Technology indeed created enormous wealth, but that wealth flowed to those who controlled the technology, not to the displaced workers.

Jensen Huang revealed a symbolic data point at the forum: six years ago, 90% of the world’s Top 500 supercomputers used CPU architecture; this year, that proportion dropped to 15%, with accelerated computing soaring from 10% to 90%. Behind this is a reallocation of hundreds of billions of dollars in computing resources, with ownership and revenue rights held by very few companies.

Fundamental Changes in the Nature of Work

Jensen Huang’s deep insight is that AI will not eliminate work but rewrite its essence. Radiologists did not lose their jobs because their core value lies in diagnosis and doctor-patient communication, not mechanical image reading—AI takes over the standardized parts, while humans retain judgment, empathy, and responsibility.

This logic also applies to other fields. As generative AI shifts from CPU to GPU architecture, the computational foundation of recommendation systems is evolving. Huang pointed out that the past 15 years belonged to the “recommendation system era,” where algorithms drove social media content streams, advertising, and recommendation logic. As this system gradually permeates into generative AI frameworks, most people will unknowingly find themselves unable to detach from these tools—and who owns the tools becomes a key issue.

Musk’s Grand Plan and the Reality

In the forum, Musk and Jensen Huang also announced cooperation to build a 500-megawatt AI data center in the Saudi desert, jointly developed by xAI, NVIDIA, and Saudi AI company Humane, with an initial investment of 50 megawatts. Meanwhile, Musk claimed Tesla will produce “truly practical” humanoid robots, with a predicted global demand of 10 to 20 billion units, aiming for Tesla to produce 1 billion units annually(, capturing over 10% of the market), with costs controlled at $10,000 and selling price at $20,000. Based on this, it is a $25 to $30 trillion industry.

Musk even predicts that within five years, the cheapest way to compute AI will be via solar-powered satellites in space—because the Earth’s received solar radiation accounts for only one in two hundred million of the Sun’s total radiation. But in reality, the U.S. generates about 460 GW of electricity annually; if AI computing requires 300 GW, it would consume two-thirds of the entire U.S. power generation, making it impossible to build enough power plants. The returns on these massive infrastructure investments will inevitably flow to the few who control computing power, models, and platforms. Ordinary workers whose productivity is increased by AI will find their bargaining power diminishing further.

The Eternal Dilemma of Scarcity

Musk claims that in the future, money will not matter, provided material abundance is achieved. But even if AI reduces production costs to near zero, scarcity will never disappear—land is limited, computing power is limited, attention is limited, and power is limited.

Human competition has never been solely for sustenance; it is more about surpassing others. If everyone can use AI to write papers, will admission rates at top universities increase? If everyone can AI-entrepreneur, will market share battles intensify? If Tesla produces 1 billion robots annually at $20,000 each, each robot must have a $20,000 purchasing power. Who can buy the first one? Who can buy ten thousand?

The future Musk envisions of universal prosperity presupposes eliminating competition itself—something that has never happened in human history.

Power Reshuffle in the AI Era

The core insight of this forum is that the beneficiaries and users of AI are not the same groups. The discussions about desert data centers, space satellites, and trillion-dollar infrastructure benefit the oligarchs who control computing power and platforms.

Musk himself admits that AI has made him busier—because his mind is full of ideas. Someone who controls computing power, models, and platforms is entirely different from someone who merely uses tools; their understanding of “work as an option” is worlds apart.

The IMF predicts that AI will impact nearly 40% of global jobs, with 70% of occupational skills changing, and developed countries being more affected—up to 60%. Technology never automatically brings equality; it only amplifies existing power structures. The future of AI is not about eliminating work but about shifting the discourse on the definition of work.

Returning to the core of the forum discussion: Musk depicts the view from the tip of the pyramid, while Jensen Huang describes the reality of the pyramid’s body. Both are not contradictory, only different perspectives. Future work will not disappear but become more fragmented, unstable, and more like a forced survival effort. For the very few who control technology, work will become a hobby; for most people, AI will only make labor an inevitable price for survival.