Zuckerberg Floats an ‘AI Air Biscuit’ at the White House
Talking Backwards
Mark Zuckerberg is getting lambasted pretty much everywhere after he appeared to be caught “groveling” to Donald Trump on Meta’s spending estimates for AI.
The gaffe occurred at a White House dinner last week attended by the biggest leaders in tech, including Zuckerberg,, OpenAI's Sam Altman and Greg Brockman, Apple's Tim Cook, and Google cofounder Sergey Brin. (With 13 billionaires in attendance and many others worth millions of dollars, the event was one of the wealthiest gatherings in the history of the White House.) The CEOs were being called upon by Trump to extoll their commitments to invest astronomical sums of money for building AI infrastructure in the United States, ostensibly to help compete against China.
Come Zuckerberg's turn, he clearly was not prepared.
"How much are you spending, would you say, over the next few years?" Trump asked.
"Oh gosh," Zuckerberg said. "Um, I mean I think it's probably going to be something, like, I don't know, at least $600 billion through '28 in the US. Yeah."
Moments later, when the formalities ended and Zuckerberg thought he was no longer being recorded, the centi-billionaire (that's someone with a net worth of $100 billion or more) was caught whispering:
"I'm sorry, I wasn't ready to do our... I wasn't sure what number you wanted to go with!"
It's a striking moment that exemplifies three things: (1) the tech industry's proclivity for pulling numbers out of nowhere, especially in our breathless age of AI; (2) all "logic and proportion have fallen sloppy dead" (from White Rabbit, released by Jefferson Airplane on September 29, 1967); and (3) there's a lot of bullshit in the bubble machine.
Artificial intelligence already ranks among the biggest investment booms in modern history. This year the large tech firms in the United States — the 33 Silicon Valley power players at Trump’s high-profile tech dinner last week — say they will spend trillions on the infrastructure needed to run AI models. OpenAI and Anthropic, the world’s leading model-makers, are raising billions every few months; their combined valuation is approaching half a trillion dollars.
The scale of these bets is so vast that it is worth asking what will happen at payback time. There’s also a conceptual gap between all the infrastructure investments to “run” AI models and the production of data to feed them. AI firms will soon exhaust most of the internet’s data, something called the “data wall” — this refers to a fundamental constraint that AI systems eventually hit due to limitations in data availability, quality, or diversity. The implications are not just economic (companies may face rapidly rising costs to get marginal improvements in AI performance), but also strategic: whoever controls the largest and most unique datasets gains a competitive advantage.
This looming bottleneck was assumed away by the Silicon Valley visionaries at the White House dinner last week. And so the Standard Talk dominates.
But even if the technology succeeds — and that’s a Big If; Wall Street's biggest fear was validated earlier this month by a new MIT study indicating that 95 percent of organizations get zero return on their AI investments — plenty of people will lose their shirts. And if it doesn’t, the economic and financial pain will be swift and severe.
Notably absent from the dinner were Elon Musk and Jensen Huang. Musk claimed he “was invited, but unfortunately could not attend,” though initial reports suggested he was not on the guest list. Huang, meanwhile, has a pattern of skipping high-profile events, preferring direct one-on-one meetings.
The dinner underscored Silicon Valley’s ‘value alignment’ problem with the Trump administration, as companies seek favorable regulatory treatment and government contracts while positioning themselves for a potential AI boom or bust. (Explore more Blue Spoon thinking here: The Delicate Art of 'Strategic Fit')
Speaking of Value Alignment
In related news, Tesla’s board this month announced an incentive package for Elon Musk that offered him a trillion dollars if he met some very ambitious goals for Tesla’s stock. Here's why that 'storyline of value' doesn't compute, via Francis Fukuyama’s Substack this morning, Our Coming Plutocracy:
"The U.S. federal budget deficit for this year is expected to come in at $1.9 trillion, and the Big Beautiful Bill is expected by the Congressional Budget Officer to add another $3-4 trillion over the next decade.
So if Musk wins this payout, he could single-handedly close a significant part of the national deficit, and personally fund all the Medicare, early childhood education, foreign aid, and other programs being cut as part of the BBB’s effort to minimize the deficit. Given that U.S. GDP last year was about $28 trillion, the payout implies that one man contributed more than 3.5 percent of the nation’s total output, while the other 340 million of us produced the remaining 96.5 percent.
Musk’s pay incentive is, frankly, ridiculous.
The idea that Musk needs this kind of reward to help his own company do well strains credibility. If Tesla’s potential failure isn’t enough to keep him focused, he probably shouldn’t be CEO in the first place. The payout is so outlandish that it’s not at all likely to happen; what is disturbing is the thinking that underlies the Tesla board’s decision."
But it’s also thinking that signals something more structurally problematic, almost endemic, to the flawed logic constraining vision and strategy to see and lead the next cycle of business and economic innovation: ‘market access’ to the middle class.
If all the value that can be already is, then market dynamics can only be about churn, conflict, accumulation and zero-sum competition — this is the current commercial orientation of the ‘control points’ in the American Way of Healthcare, including United Healthcare “Spending Big on Trump Allies to Fix Its Washington Problems” and Express Scripts telling attendees at Morgan Stanley’s Global Healthcare Conference last week that it’s doing what’s necessary to ensure its “three value creation levers aren’t compromised."
Static or contracting economies make people cruel and shortsighted.
I would argue we need a palate cleansing, a broadening of the horizons, better strategy stories to break the deep freeze of convention sustaining the status quo. Because if the agenda is imagination to compete in the age of computing, Grace Slick gave us the right diagnosis of The Problem almost 60 years ago: the White Knight is talking backwards.
/ jgs
John G. Singer is Executive Director of Blue Spoon, the global leader in positioning strategy at a system level. Blue Spoon specializes in constructing new industry narratives.