ChinaTalk's Jordan Schneider: US is making every mistake in the playbook to lose the tech war with China

Summary

Jordan Schneider, founder of ChinaTalk, argues the US is systematically dismantling the advantages that made it a near-certain winner in the tech competition with China — and doing so voluntarily.

The brain drain problem

Schneider's sharpest concern is the visa crackdown on STEM talent. He cites 40 visa revocations already, affecting biotech and computer science PhDs, and draws a direct historical parallel to Qian Xuesen — a Chinese-born MIT graduate who worked on classified US missile programs before McCarthy-era persecution sent him home, where he built China's nuclear and ICBM programs. Schneider's thesis is that a handful of researchers and founders will determine the trajectory of technology over the next 30 years, and the US is needlessly telling those people they aren't welcome.

The chip export question

The Trump administration has allowed Nvidia to export H20 inference chips to China — a sale Biden had blocked. Schneider puts the revenue figure at roughly $2 billion and argues Nvidia doesn't need it. The steel-man case for allowing the sales is that $2 billion in AI compute can't build a GPT-5-class model and is therefore strategically irrelevant. Schneider is skeptical: the asymmetric downside risk — that those chips enable fast-following on frontier US models — isn't worth the upside. The hyper-scalers are each spending roughly $65 billion in capex this year; a $2 billion Chinese inference cluster isn't going to close that gap, but selling it directly into China removes the option to cut off access later. The current workaround — routing chip sales through Malaysia and Singapore — at least preserves that lever.

TikTok, DJI, and strategic incoherence

While the US imposes tariffs on Chinese goods, it has failed to ban TikTok or restrict DJI drones, both of which Schneider views as obvious national security calls. China, he argues, is laughing its way to the bank — extracting real intelligence and influence value from US consumers while the US focuses on tariffs on cheap manufactured goods. Ukraine still buys Chinese drone components because there's no viable ex-China supply chain, which illustrates the gap between where US industrial policy should be aimed and where it actually is.

Trade war architecture

The 2019 round of US-China tariffs was manageable partly because the US wasn't simultaneously fighting trade wars with Canada, the UK, Australia, Vietnam, and the rest of its allies. The 76% tariff on Vietnam — later rolled back — is a symptom of an approach Schneider describes as tactically incoherent. The real risk isn't just that China benefits directly; it's that the rest of the world builds a trade ecosystem that excludes the US, and American exporters and consumers are locked out of it. The US represents roughly 25% of global GDP, which means 75% of the world is calculating who else to trade with.

China's commercial aviation gap

China has blocked Boeing orders as a retaliatory move. COMAC, China's state-owned aircraft manufacturer, has spent decades trying to build Boeing and Airbus competitors with limited success. Schneider attributes the slow progress to state ownership and weak incentive structures — the same dynamics that produced world-beating EV companies when private capital and phone manufacturers flooded the auto sector never took hold in aviation. If Chinese airlines can't fly COMAC, orders go to Airbus, which hurts the US but doesn't solve China's problem.

The AI wild card

Schneider had viewed AI dominance as the US's strategic escape hatch — a lead large enough to offset policy missteps elsewhere. That confidence is eroding. Selling H20 chips into China, destabilizing the US capital base that funds AI expansion, and scrapping the AI diffusion rules so that chips flow to Saudi Arabia, Qatar, and Brunei rather than staying in allied hands all eat into that lead. The US might still pull it out, but it's taking on avoidable risk across every front simultaneously.