The prophet margin – when CEO spit-balling nudges just close enough to AI strategy


It is fast becoming the rule that when a high-profile tech CEO makes a bold yet untestable announcement, you should always look elsewhere for a reason. Witness Tesla, SpaceX, and X supremo Elon Musk’s claim last November that, in the future, robots will be an “infinite money glitch” that will, by some unspecified means, solve global poverty and make jobs into an optional pastime for humans. 

Yes, none of us will have jobs except for fun, it seems, yet we will all mysteriously be able to afford a six-foot humanoid called Optimus and a monthly AI subscription – food, heat, water, and a place to live be damned. 

By happy chance, that claim – unverifiable in the real world – coincided with an investor call in which Musk was seeking approval for a pay package that could be worth up to $1 trillion over a decade. The deal was approved, despite the fears expressed by some shareholders about Tesla’s brand taking a nosedive after months of DOGE controversy and an ill-advised (ahem) ‘Roman salute’ by Musk onstage.

Job done. We’re a robot company now, their leader suggested. Cars? They’re so last century – even though Tesla’s strategy of pitching Optimus at the unproven, high-risk domestic market remains questionable at best.

Fast forward to this year, and Musk repeated the trick with a suggestion published on the SpaceX website that the company could – hypothetically, mind – put a million AI data centers in orbit. That would be seventy times the current number of all satellites, as the first stage in Earth becoming a Kardashev Stage Two civilisation, one capable of harnessing all its star’s energy, he explained. 

That the comment stopped short of being a strategy announcement is important: it created plausible deniability. Yet however preposterous and untestable an idea might seem, social platforms enable leaders to think aloud to an audience of millions. And so, such a statement gets amplified to the markets as well as to followers and acolytes, and it scales to global news. 

In short, it’s tantamount to a strategy statement in its effect, but without being a firm commitment that auditors and regulators can ever test. “Hey, I was just spit-balling, guys,” a CEO might say. “I was just thinking aloud. I wasn’t being serious.” 

As a doubtless unrelated side effect, such prophetic pronouncements do, of course,  grab headlines away from stories like the Grok chatbot being used to make millions of non-consenting images of women, and illegal ones of children, for example, or other uncomfortable topics. Suddenly, the world and its markets have a new distraction: a million data centers in space! Ker-ching!

Not just Musk...


All this talk of untestable claims and CEO spit-balling brings us, unexpectedly, to NVIDIA CEO Jensen Huang. If you were thinking that the days of tech leaders dangling AGI (Artificial General Intelligence) as 'investor bait' were over, you would (sadly) be wrong. 

First, a quick recap. In his NVIDIA GTC keynote in San Jose last week, Huang made 18 product announcements and suggested that, in the future, the world’s employees would have token budgets, and we would all need a “claws” (autonomous agent) strategy as OpenClaw becomes the ChatGPT of agentic AI. In short, humans’ main role in the future will be providing a revenue stream for AI companies – even if we don’t consent.

Cue an instant shares uplift, right? Nope. Alas NVIDIA’s shares fell during his epic speech, with war and energy security being top of mind for the markets rather than yet more power-guzzling racks and robots.

But Huang wasn’t done with the media yet – not by a long way. After all, being the leader of a multi-trillion-dollar corporation demands living in a glass box in public every day, like a David Blaine made of money. 

Not content with having failed to inspire the markets last week, Huang this week opted to invoke OpenAI’s strategy of talking up AGI instead – the thing that OpenAI's Sam Altman said was no longer “super-useful” last year after the damp squib of GPT-5. Speaking on the Lex Fridman podcast on 23 March, Huang said of the timing of AGI's arrival:

I think it’s now. I think we’ve achieved AGI. 

Such statements are never made by accident. Dutifully, the mainstream media parrotted, “We have reached AGI, says Huang”, without questioning the claim – which he made about the industry as a whole – or engaging their critical faculties. The headline was everything, and the details were completely irrelevant: today, almost no one reads beyond the tweet.

As if to prove the point, NVIDIA’s shares moved up 1.5 percent on the day, reversing the decline caused by last week’s detailed statement of product, vision, and strategy. Peak 2026!

But unlike OpenAI’s Altman, Huang’s claim began to unravel almost as soon as he said it. (With Altman, it usually takes a few weeks.)

Asked if AGI could run a company, Huang said it was “possible”. But then rapidly explained that the chances of an autonomous AI agent building and running an empire like NVIDIA were “zero percent”. Ah yes, agents might be able to get their claws into your company or mine, but not NVIDIA, no chance. Not even single digits of possibility!

But Huang needs you to believe that you can’t automate the prophet margin, but you can automate everything else. You just need faith in the self-appointed seer schtick, and everything else will fall into place, it seems.

But the question is, why did NVIDIA’s Huang feel the need to restart the AGI hype machine, and grab the headlines with such apparent ease? After all, isn’t he presiding over the world’s most valuable company by far – one that is worth $600 million more than Apple, and which made a profit of $120 billion last year? Surely the man can take a week off?! Of course, NVIDIA’s stock is still down $14 since January, so maybe every little helps as the saying goes?

My take

Back at OpenAI, Altman told Forbes last month:

We basically have built AGI, or are very close to it.

It's yet another untestable, meaningless CEO statement, which he later added was meant in a “spiritual” sense, not a literal one. 

TV mogul  Logan Roy in Succession once said of his children, "None of you are serious people". I don't know why that springs to mind right now. Is anyone even trying to say  this stuff with a straight face anymore? Or are we all just being trolled for clicks and trillions?

The problem for these social, stock, and brand-focused leaders in 2026 is this: for buyers, all this hype and empty wordplay mean ever-diminishing returns. We urgently need them to do better than just keep their own investors and followers happy. We need real insight plus measurable results and a demonstrable ROI. 

The hype machine is running out of hot air, fast.