“Every Chip Is Getting Used Instantly” — Here’s Why Google’s AI Dominance May Be Unstoppable

I’ve been following Peter Diamandis’s commentary on AI infrastructure for the better part of two years now, and his recent observation on the “Moonshots with Peter Diamandis” podcast (EP 246) is stunningly accurate.

His co-host laid it out: “every chip that’s coming out is getting used instantaneously. There is not an idle memory or processing chip anywhere in the country.”

In that context of total scarcity, who owns the chips owns the future.

According to Diamandis, the answer is overwhelmingly Google. A chart referenced in the episode shows Google’s chip holdings exceed those of entire countries, including China, followed by Microsoft, Amazon, Oracle, and XAI. Diamandis called Alphabet (NASDAQ:GOOGL | GOOGL Price Prediction) “the dominant force” in AI and said it “will win in the long run.”

Larry Page Saw It in 2016

The foresight story starts a decade ago. Larry Page gets primary credit for this advantage, having seen the AI chip opportunity in 2016 “before anyone was thinking about this stuff.” Google began developing its Tensor Processing Units internally around 2015-2016, giving it a proprietary silicon foundation that no competitor has fully replicated.

The numbers validate that bet. GOOGL has returned 730% over the past ten years, from $38.68 in April 2016 to $321.31 today. A decade of compounding proprietary infrastructure built that return.

Now Alphabet is doubling down. Sundar Pichai confirmed 2026 CapEx guidance of $175 to $185 billion, nearly double FY2025’s $91.4 billion. Google Cloud already runs at an annual run rate exceeding $70 billion, with Q4 2025 cloud revenue up 48% year-over-year.

The Competitive Gap Is Growing

Microsoft (NASDAQ:MSFT) is spending aggressively too, with Q2 FY2026 CapEx of $29.88 billion, nearly doubling year-over-year, but it remains dependent on NVIDIA and its OpenAI partnership. Amazon (NASDAQ:AMZN) is building its own silicon with Trainium and Graviton, with a combined annual run rate over $10 billion growing triple digits, and plans roughly $200 billion in 2026 CapEx. Oracle (NYSE:ORCL) plays a “chip neutrality” strategy, running all top five AI models in Oracle Cloud rather than owning the silicon itself.

The Monopoly Question Few Are Asking Loudly

Diamandis is askin the right question: “When are they going to run into monopoly concerns?” His answer was geopolitical. The current administration’s “beat China at all costs” posture means antitrust pressure on Google’s chip dominance is unlikely in the near term. When national security and competitive advantage point the same direction, regulators tend to stand aside.

If you believe the AI infrastructure race is won at the chip layer, and that Google’s decade-long head start compounds with each passing quarter of $175 billion in new spending, the Diamandis thesis carries real weight. The risk is that scale alone does not guarantee model quality or enterprise adoption. But owning the picks and shovels of an arms race, at a scale that rivals nation-states, is a position most investors would take.