A viral AI meme may have accidentally summarized the power structure of the entire industry.

The X post from @iam_smx jokes about “Lord Elon Musk,” “Scam Altman,” and Dario Amodei all showing up to request GPUs from “King Jensen.” The wording is obviously satire, but the underlying economics are real: some of the most powerful AI executives on Earth still need access to NVIDIA’s compute stack.

Lord Elon Musk, Scam Altman, and Dario requesting GPUs from King Jensen is absolutely hilarious 😂pic.twitter.com/0BwM8hHEdo

— SMX 🇺🇸 (@iam_smx) October 5, 2026

The Joke Works Because NVIDIA Really Is the Court

At CES 2026, NVIDIA’s own Rubin platform announcement carried endorsements from all three camps. Sam Altman said intelligence scales with compute. Dario Amodei said NVIDIA’s Rubin efficiency gains enable longer memory, better reasoning and more reliable outputs. Elon Musk called Rubin a “rocket engine for AI” and described NVIDIA as the gold standard for frontier-scale training and deployment.

That is an extraordinary alignment. OpenAI, Anthropic and xAI compete for users, researchers, capital and cultural influence, but all three can still converge on the same supplier when the conversation turns to high-end accelerators, networking and full-stack AI infrastructure.

OpenAI: Millions of GPUs and Gigawatts of Compute

OpenAI’s dependence is measured at infrastructure scale. BitcoinVersus.Tech recently covered OpenAI locking 8 GW into an Ohio AI campus buildout, part of a broader race to secure land, power, data-center shells and accelerators before demand outruns supply.

The meme’s “asking the king” framing is funny precisely because the order sizes are no longer about racks of GPUs. Frontier labs are effectively negotiating in gigawatts.

Anthropic Is Spending Like Compute Is Sovereign Territory

Anthropic is no exception. BitcoinVersus.Tech reported that Anthropic has locked in roughly $518 billion of long-term AI infrastructure commitments. Those obligations span cloud providers, chips and hosting capacity because the company cannot assume frontier compute will simply be available when needed.

That backdrop also makes Jensen Huang’s position unusual. NVIDIA is not just selling one component into the boom; it sits across accelerators, NVLink, networking, systems software and increasingly the design of entire AI factories.

xAI Can Build Fast, but It Still Needs the Silicon

Elon Musk’s companies have pushed unusually fast on physical buildout. BitcoinVersus.Tech just covered SpaceXAI doubling its “Minihard” supercompute effort in Memphis. But even the fastest data-center deployment team still needs the accelerator supply chain underneath it.

That is why the comedy lands: Musk can own rockets, Altman can raise historic amounts of capital, and Amodei can sign hundreds of billions of dollars in infrastructure commitments, but none of that eliminates the question of who can deliver the most capable compute at the right scale and time.

Jensen’s Real Power Is Being the Common Denominator

TechCrunch reported in September that Huang described NVIDIA as foundational to the AI ecosystem and said the company tracks essentially every gigawatt of land, power and data-center shell tied to its partners. That visibility gives NVIDIA a perspective few individual AI labs can match.

The result is almost feudal in the funniest possible way. The AI labs fight each other for models, talent, users and prestige. Then they all arrive at the same castle asking when the next generation of GPUs ships.

So yes, “King Jensen” is a meme. But underneath the crown is one of the most strategically powerful positions in modern technology: being the supplier that your rivals all need.