The AI Boom is Real and This is How We’re Approaching It

Here’s a bet: by the time you finish this sentence, someone in Washington will have said the words “AI bubble” without ever having deployed a model in production.

This past June gave the crowd some ammunition. Or so it looked. On June 23, the tech-heavy Nasdaq dropped 2.2% and the S&P fell 1.4%, while South Korea's KOSPI collapsed 10% and briefly halted trading to prevent a crash, and shares of memory-chip giants Samsung and SK Hynix lost 12% in a single morning. Micron Technology, whose stock had risen nearly 800% over the previous year on AI-memory demand, plunged 13% amid that sell-off. Add in a Federal Reserve that has formally identified AI as one of the top systemic risks to U.S. financial stability, plus roughly $725 billion in projected Big Tech AI capex against a Gartner forecast of $2.53 trillion in total AI spending worldwide this year, and the bubble case starts writing itself.

They’re wrong. Not because the spending is safe – it isn’t. Not because every company burning cash on inference survives the next two years – most won’t. They’re wrong because they’ve mistaken a market correction for a technological one and those are not the same animal.

In 1867, the United States of America bought Alaska for two cents an acre and the press had a name for it: Seward’s Folly. An icebox. A vanity purchase by a Secretary of State nobody much liked. Thirty years later, gold turned up in the Klondike. Sixty years after that, so did oil. Nobody remembers the punchline. They remember the acreage.

There’s a name for what the bears think is happening: the greater fool theory. The bet isn’t that the asset is worth the price – it’s that someone later, dumber, or more desperate will pay more for it before the music stops. Award-winning screenwriter Aaron Sorkin liked the idea enough to build a full hour of “The Newsroom” around it. It’s the same logic that sent tulip bulb prices in Holland soaring to the price of a house in the 1630s, right before the market for them evaporated overnight – a fine description of an asset with no floor once buyers stop showing up. It struggles once the asset in question is a data center bolted to the ground in Ohio with a power purchase agreement attached to it. You can’t unload concrete and turbines onto the next fool the way you dump a stock ticker or a bulb. The floor under this boom is physical capacity training the next round of frontier models – not sentiment waiting for a bigger sucker.

The transcontinental railroad looked like a folly. Rural electrification looked like a folly. The interstate highway system looked like a folly, right up until the country reorganized itself around it and stopped noticing. Generational infrastructure always reads as excess to the generation writing the check. AI is next on that list and the analysts measuring a twenty-year bet against a twelve-month ledger are making the same mistake the newspapers made in 1867.

This is also, not incidentally, why the current Administration has staked so much of its opening years on getting out of the way. Last July, the White House released “Winning the Race,” its AI Action Plan – more than 90 federal policy actions organized around three pillars: innovation, infrastructure, and international security. Strip the branding and the substance holds up:

• Deregulation First – Fewer permitting bottlenecks for data centers and the power generation that feeds them.
• Chips as Strategy, Not Commerce – Export controls treated as a national security lever against Beijing, not a trade dispute.
• Procurement as a Market Signal – Federal contracts steered toward AI developers whose systems are judged free of top-down ideological override.
• Codification, Not Just Executive Order – the House Foreign Affairs Committee has said it intends to codify most of President Donald J. Trump’s plan into law in the months ahead.

Say what you want about the particulars – we have opinions, some sharper than others – but the through-line is coherent. The Administration’s own framing, offered under oath before Congress this past January, was that the country should meet this moment with “responsible boldness” rather than fear. We support that posture. Not uncritically – bureaucracies rarely deserve uncritical support, ours included – but directionally. A country that regulates a technology into stagnation while its chief rival subsidizes one into dominance has already lost the argument it thinks it’s still having.

That’s the outside view. Here’s Connector’s, because a firm that spends its days telling other people how to communicate has an obligation to be honest about how it communicates.

The ideas, arguments, theories, opinions, and conclusions you read here are ours. Every piece begins with the experience, perspective, judgment, and – perhaps most importantly – the point of view of its contributor. When AI is used in our publishing process, it serves as a tool, not an author. It may assist with editing, grammar, syntax, formatting, organization, or other elements of the editorial process. It does not replace the contributor’s thinking, experience, judgment, or point of view.

That’s not a disclaimer. It’s a description of a workflow we’d run whether anyone asked or not.

One more thing, while we’re being honest. Somewhere in the last several years, the em-dash became a suspect – a stylistic tell supposedly giving away the machine behind the writing. So did the ellipsis. Neither claim survives contact with anyone who’s read this newsletter for more than a week or sat across from me writing for the last twenty years. I use ellipses because that’s how the thought actually moves . . . a pause built into the sentence, not sprinkled on top of it by a chatbot with a style guide. The em-dash is the same story – a tool for pace, not a signature stamp. Assume what you like. We were writing this way before the tools existed to imitate it.

So: the boom is real. The tool is useful. The judgment stays ours.

Rob Burgess is a national Republican strategist and Chief Executive Officer at Connector, Inc. – a boutique government relations, public affairs, and political strategy firm with offices in Washington, D.C.