On ToTheMoon, Jensen Huang is not a subject of NVIDIA news so much as a supplier of formulas the hosts use to describe the economics of AI. His quotes surface around the price of compute, how many tokens an employee ought to burn, open models, and how to compete with China. The hosts keep his interest in view: in 2024 they noted that Huang is optimistic because AI makes him a great deal of money, so it was telling that even he conceded hallucinations would not be solved quickly (episode 035).
In 2026 his statements became a regular point of argument. His claim in an interview with Lex Fridman that AGI has broadly been achieved was dismissed bluntly (episode 103). His remark that AI cannot build an NVIDIA but can build a decent company was cited as a sober estimate (episode 110). The formula that stuck was the token economy: a programmer earning half a million dollars a year who does not spend as much on tokens is a bad sign. The channel invoked it when discussing the cost of access to the strongest models (episode 139) and in a review of the new race in which, per Huang, everyone will be selling tokens (episode 161).
The second thread is openness and China. The hosts set Huang's line, that the United States should not wall itself off from China and should keep supplying American hardware there, against Amodei's position at Anthropic (episodes 127 and 152). His advocacy of open models (episode 150) resurfaced in the breakdown of Zuckerberg's manifesto, backing, with Mira Murati's stance, the argument that open weights will gain share simply on price (episode 152).
Author's trends
the author's thoughts whose episodes feature Jensen Huang · blue marks the months they meet