Jensen Huang, CEO of Nvidia, has forcefully pushed back against predictions that artificial intelligence will lead to widespread job destruction, calling such fears “exactly backward.” Speaking at Y Combinator’s Startup School, Huang argued that AI automates specific tasks rather than entire jobs, allowing workers to focus on higher-value activities. He cited radiologists—long predicted to become obsolete—whose demand has increased because AI accelerates scan analysis. Similarly, software engineers now use AI to write code faster, enabling companies to tackle larger backlogs and hire more developers. Huang dismissed Anthropic CEO Dario Amodei’s claim that AI could erase 50% of entry-level white-collar jobs as “complete nonsense.”
Huang’s optimism is supported by the PwC 2026 Global AI Jobs Barometer, which analyzed over one billion job ads and found that roles requiring specialized AI skills are growing nearly eight times faster than the overall market. Those positions command a 62% wage premium. However, the job market for recent graduates tells a different story: unemployment among young graduates hit 5.6% in early 2026, and a Stanford Institute for Economic Policy Research brief points to AI as a possible factor, as many entry-level tasks are precisely the kind AI automates. The New York Federal Reserve’s data shows that unemployment for all college graduates aged 22–27 was 5.6% in March 2026, compared to 3.1% for all graduates, while Uber recently cut 10% of its customer service staff explicitly citing AI.
Separately, deVere Group CEO Nigel Green has issued a stark warning about the financing behind the AI boom. He told investors to look beyond revenue headlines and examine whether infrastructure spending is becoming a circular flow among the same handful of players. Green highlighted the recent semiconductor sell-off in Seoul and Tokyo, where Samsung Electronics and SK Hynix lost nearly a tenth of their value in a single session, as a sign of fragile confidence. He cautioned that capital is “recycling” within the ecosystem: Nvidia funds OpenAI, which pays Oracle and Microsoft for cloud capacity, who then buy Nvidia chips. “The same dollar gets counted as revenue three times on its way around the loop. This is not demand. It’s an accounting trick wearing a growth story as a costume,” Green said.
Green also warned that Chinese AI developers are closing the technology gap rapidly, shipping frontier open-source models trained on domestic chips at lower cost. He believes lower-cost Chinese AI could pressure the economics of today’s infrastructure investments. While he does not advise abandoning AI exposure, Green urges investors to scrutinize balance sheets, customer revenue sustainability, and competitive resilience. “The companies that win will be the ones with real customers paying real money, margins that don’t depend on their own supplier lending them the cash to buy the product, and tech that still works if the financing dries up,” he said.
