Nvidia (NVDA) CEO Jensen Huang is challenging the people warning that AI could end humanity, suggesting their alarming predictions may serve another purpose.
Speaking with CBS News, Huang rejected the ‘AI apocalypse’ debate and questioned the motives behind those spreading that fear, adding a sharper edge to an increasingly public industry split.
OpenAI and Anthropic leaders, alongside Elon Musk, have backed calls to slow advanced AI development over safety concerns and fears that the technology might be advancing quicker than companies can control it, as reported by Financial Times.
For Nvidia investors, that debate carries financial weight. Shares have returned 20% year-to-date and 23% over the six months, according to Seeking Alpha data.
A coordinated slowdown could complicate the spending boom supporting its chip business. Huang’s response, however, reaches beyond defending continued investment.
Jensen Huang questions the motives behind AI doomsday warnings
Huang believes that the people predicting an AI apocalypse may have reasons beyond protecting the public. In his CBS interview, Nvidia’s CEO questioned their motives while rejecting a specific prediction of catastrophe by 2030.
More Nvidia:
- Nvidia just made a move Wall Street wasn’t ready for
- Nvidia just locked down deal that changes AI race
- Nvidia stock is doing something it hasn’t done in years
“I believe the claims of the end of the world, stirring fear across America, and doing it by the people who are doing it make no sense to me,” Huang said.
“So they must be doing it for ulterior reasons.”
That moves the argument beyond whether powerful AI needs safeguards. Huang is questioning why people involved in developing the technology are presenting its future in such alarming terms.
His explanations were speculative, however. He offered possible motives without establishing which, if any, applied.
“Maybe it’s political, maybe it’s otherwise, maybe it’s just attention grabbing,” he said, adding, “I don’t know what the reason is for it.”
On the prediction itself, Huang left little room for interpretation.
“2030 is not going to be the end of the world,” he said. “There is 0% chance that’s going to be the end of the world.”
That certainty has a specific scope, though. It addresses an end-of-the-world scenario by 2030, which leaves other questions of cyberattacks, disruptive failures and control over increasingly capable systems unresolved.
His challenge is directed at the doomsday narrative and the people promoting it.
Bloomberg / Getty Images
AI’s everyday failures complicate the doomsday debate
That said, Huang declined to speak for President Donald Trump when pressed on whether those fears were a “hoax”.
“Well, I don’t want to speak for him,” Huang told CBS, before laying out his own objections, which underscore his willingness to challenge catastrophic predictions, without at the same time establishing that every concern about AI is baseless.
Google Brain cofounder Andrew Ng has also challenged the hype surrounding AI with broadly human-level capabilities.
According to Ng, AI extinction warnings were “much more science fiction than science,” according to Startup Fortune. Ng argues that renewed doomsday rhetoric is mainly a public-relations effort to shape regulation, while testing, adoption of safeguards and continued research offers a better path to making AI safer.
Recent evidence highlights more immediate shortcomings.
September research from financial technology firm Saturn tested 18 models against 121 financial questions, repeated five times. Across more than 10,000 responses, 57% failed Saturn’s criteria, rising to 88% on harder questions. Failures included factual errors, missing information and omitted warnings, as reported by Financial Reporter,
These reports present a practical problem in that impressive answers can still be unreliable. That does not remove future catastrophic risk. It does give businesses and investors a concrete test today in whether AI works reliably enough to trust.
Nvidia’s valuation puts the burden on earnings delivery
For Nvidia investors, the big issue is whether AI spending can deliver the earnings already embedded in expectations. Huang’s confidence offers reassurance, but the stock’s valuation rests on execution.
Seeking Alpha data puts Nvidia at 23.57 times forward non-GAAP earnings, roughly 45% below its five-year average and just 5% above the sector median. That suggests investors are paying a relatively modest earnings premium for substantial projected growth.
The estimates explain why.
Analysts expect earnings per share of $9.31 for the fiscal year ending January 2027, rising 68.5% to $15.68 the following year. On that later estimate, the stock trades at 14.17 times earnings.
That creates an attractive possibility where Nvidia could grow into its valuation without needing investors to assign it a richer multiple.
But those estimates also carry the investment risk. The January 2028 low EPS estimate is $9.80, far below consensus, showing how widely expectations diverge. Meanwhile, forward enterprise value to sales remains 12.81 times, versus 3.42 for the sector.
In other words, the earnings valuation depends a ton on Nvidia sustaining strong profitability as revenue expands.
Investors should watch customer spending commitments, deployment schedules and earnings revisions. A coordinated AI slowdown could pressure those assumptions. Continued delivery would strengthen the valuation case far more than any rebuttal of doomsday predictions.
Related: Palantir CEO offers surprising solution to AI threats that involves Washington, D.C.
