Episode summary
Nvidia’s $750 Billion in Deals Reignite Circular AI Fears | Bloomberg Intelligence
Bloomberg Intelligence
· · 17 min · Hosts: Paul Sweeney, Scarlet Fu · Guests: Ed Ludlow, Laura Martin, Shawn Donnan
Summary
Nvidia’s latest AI infrastructure announcements revived the debate over whether the company is creating circular financing structures to inflate demand. A guest said the reported SK Group package is best understood as a multi-year bundle of chip purchases, data-center spending and co-investment, but said the OpenAI arrangement is more explicit because Nvidia is both guaranteeing and financing deals that ultimately buy Nvidia technology. He argued the market now wants proof that rising capex creates measurable usage and revenue, not just bigger spending plans.
The show then turned to Meta, where Laura Martin argued that the company’s employee culture and values are unusually weak relative to other large-cap tech names. She said Meta pays far more in stock-based compensation than peers because employee ratings are poor, strategic pivots are constant, and the result is higher churn among the best workers. In her view, that makes execution harder and dilutes shareholder value.
Finally, Shawn Donnan used the tin can industry as a warning about Trump-era steel tariffs. He said the tariffs have raised costs, failed to restore meaningful domestic production, and left the U.S. more dependent on imported steel. The story’s broader argument was that protectionist promises of jobs and investment have not been matched by results.
Key takeaways
- A guest said Nvidia’s SK Group deal is a multi-year package of chip sales, data-center spending and co-investment, which helps explain why the numbers look circular.
- The guest said Nvidia’s OpenAI financing is more clearly circular because Nvidia is guaranteeing and backing stoping deals that still end up buying Nvidia technology.
- Laura Martin said Meta’s employee ratings are 2.8 out of 5, the lowest among the 23 stocks she covers, and argued that weak culture drives higher compensation costs.
- Martin argued Meta’s rapid strategic pivots, from the metaverse to generative AI to chips, make execution harder and increase employee churn.
- Shawn Donnan said the tin can industry shows tariffs raising prices without delivering the promised surge in U.S. manufacturing or jobs.
Notable quotes
“The biggest technology companies had to say that they were going to spend more every year. Capex would go up and they needed to show topline growth”
“In the open AI case, what we're reporting is much more explicit where, you know, Nvidia's both both guaranteeing and backs stopping but also providing financing”
“their ratings by employees today is a 2.8 out of five, which is the lowest of all the companies we cover by a lot”
“I think it means you have higher employee churn and the people who churn are the people who can get the next job which means they're good people”
How the episode opens
Bloomberg Audio Studios, podcasts, radio, news. You're listening to the Bloomberg Intelligence Podcast. Catch us live weekdays at 10 a.m. Eastern on Apple CarPlay and Android Auto with the Bloomberg Business App. — Listen on demand wherever you get your podcasts or watch us live on YouTube. Nvidia is working on a fresh round of AI infrastructure deals potentially worth more than $750 billion, including an AI initiative with SKH Highex for almost half a trillion dollars. The CEO of Nvidia, Jensen Huang, spoke in an exclusive interview with Ed Ledllo about investing in South Korea. Take a listen.
This is the golden ages for Korea. As you know, their semiconductor business is booming, their industrial business is booming. You know, this is a country that has the ability to help the world build out the AI infrastructure that they're incredibly adept at uh adopting new technologies and is a really technologically uh forward leaning society.
All right, that was Nvidia's CEO Jensen Wong. And joining us now is BTE anchor uh Ed Lello. Ed, the numbers just keep coming. The investments just keep coming. The circular deals keep coming. What's the view of of that type of structure as this AI infrastructure evolves? How's what's the feeling out in Silicon Valley about how this is evolving?
Part of the reason there is even a debate about it being circular or not is that there isn't a great explanation of what the numbers actually represent. So you've done the reporting about, you know, what Dean has written on the the terminal on OpenAI. If we just take SK as a case study, it's $500 billion or half a trillion dollars.
And at some point in that conversation with Jensen, I just said, "Could you just explain what that number actually is?
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- circular ai deals
- openai computing lease
- meta culture and talent costs
- china cxmt ipo
- steel tariffs tin can industry
Companies and tickers mentioned
- NVDA
- META
- Nvidia
- SK Group
- OpenAI
- Meta
- Needham & Company
- CXMT Corp.
- Glassdoor
Watch the episode
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from Bloomberg Intelligence
-
Anthropic’s IPO Is Said to Be About Six Times Oversubscribed | Bloomberg Intelligence -
Instant Reaction: Nvidia CEO Jensen Huang Ousted by Nvidia Board | Bloomberg Intelligence -
Nvidia’s Huang Ousted by Board in Move That Roils AI Economy | Bloomberg Intelligence -
CoreWeave Shares Soar After AI Computing Firm Eases Debt Fears | Bloomberg Intelligence -
Anthropic Targets $1.4 Trillion Valuation in Blockbuster IPO | Bloomberg Intelligence -
Nvidia Probe Into CEO Affair Raises Thorny Questions for AI | Bloomberg Intelligence
Bloomberg Intelligence, this episode and its artwork belong to Bloomberg. Octagon is not affiliated with the show; this page is an independent summary.
