Jersey Mike's IPO illustrates how bad the AI hype has become
There's a moment in every bubble when the absurdity becomes undeniable. For the dot-com era, it was Pets.com burning through $300 million to sell dog food online. For crypto, it was Long Island Iced Tea rebranding to "Long Blockchain" and watching its stock triple overnight. For the current AI mania, that moment arrived when a submarine sandwich chain — the kind where Danny DeVito shouts about "authentic" ingredients in TV spots — felt compelled to sprinkle "artificial intelligence" across its IPO filing like fairy dust.
Twenty-two times. That's how often "artificial intelligence" or "AI" appears in Jersey Mike's S-1 registration statement. A company whose core competency involves slicing cold cuts, baking bread, and managing franchise relationships. A business model fundamentally unchanged since the first location opened in Point Pleasant, New Jersey, in 1956. Yet somehow, investors are expected to believe that large language models or neural networks represent a material factor in the future of the Italian sub.
This isn't just window dressing. It's a diagnostic symptom of a market that has lost all grounding in reality.
The contagion has escaped the tech sector
We've watched this progression before. First, the genuine AI companies — your OpenAIs, your Anthropics — attract capital. Then the adjacent infrastructure plays: Nvidia, the data center builders, the networking gear makers. Then the "AI-enabled" software startups, eachclaiming a proprietary model that turns out to be a wrapper around GPT-4. Then the non-tech companies desperate for a multiple expansion: consultants, retailers, manufacturers all suddenly "leveraging AI."
But a sandwich franchise? This is the "Long Blockchain" phase. This is the moment where the hype has completely untethered from any plausible value creation.
Jersey Mike's isn't even trying to sell you AI. They're not pitching a proprietary kitchen robot that perfectly portions provolone. They're not claiming a computer vision system that detects stale bread before it reaches the customer. The filing admits, with remarkable candor, that "We are beginning to use AI Technologies in our business" — a sentence that means precisely nothing and appears designed solely to trigger algorithmic buyers and lazy analysts scanning for keywords.
Boilerplate as disclosure theater
The most telling detail is the risk factor section. Jersey Mike's includes AI-related risk warnings — standard boilerplate about data privacy, bias, regulatory uncertainty, hallucinations. But they don't specify what AI systems they're actually deploying that could generate these risks. Because there almost certainly aren't any. The risk factors aren't there because the company has evaluated its AI exposure; they're there because the lawyers know that omitting them would look suspicious in 2024.
This is disclosure theater. Performative compliance. The same instinct that drove companies in 2021 to add "metaverse strategies" to earnings calls, or in 2017 to announce "blockchain initiatives" that were press releases and nothing more.
Consider the contrast: the word "weather" appears five times in the filing. "Lightning" appears zero times. Yet a Jersey Mike's franchise in Texas was literally struck by lightning in 2021. A genuine, physical risk to assets and operations — ignored. Meanwhile, hypothetical AI risks that almost certainly don't exist for this business merit multiple paragraphs.
That inversion tells you everything about where capital attention sits.
The Bending Spoons precedent
Jersey Mike's isn't alone. The source reporting notes Bending Spoons, the Italian tech conglomerate that buys decaying software assets — Evernote, Meetup, Issuu — and filed for its own public offering. Their pitch? They're not a roll-up play. They're an "AI company" now. Never mind that their core business is acquiring distressed properties and cutting costs. The AI narrative is the only one that commands a premium multiple in public markets.
This pattern repeats across venture capital. Founders raising Series A rounds for B2B SaaS tools, logistics platforms, healthcare admin software — all feel pressure to articulate an "AI strategy" whether or not machine learning improves their product. The capital allocators have signaled clearly: no AI story, no premium valuation. So everyone manufactures one.
History doesn't repeat, but it rhymes violently
We know how this ends. The dot-com bubble left behind Amazon, Google, and eBay — real companies with real moats — but vaporized trillions in speculative excess. The crypto winter purged countless tokens but left Bitcoin, Ethereum, and legitimate DeFi infrastructure. The AI bubble will similarly leave behind transformative technology: foundation models, coding assistants, scientific research acceleration, medical diagnostics. The technology is real. The valuations, the narratives, the forced insertions into sandwich shop IPO filings — those are the froth.
The tragedy is that the froth obscures the signal. Legitimate AI companies building hard tech struggle to differentiate from the noise. Investors allocate capital based on keyword frequency rather than technical moats. Management teams waste cycles crafting AI narratives instead of improving products.
The sub is still just a sub
Jersey Mike's makes a perfectly fine sandwich. Their franchise model works. They have 3,000+ locations and a loyal customer base. That's a decent business worth a reasonable multiple. But it's not an AI company. It will never be an AI company. No amount of "beginning to use AI Technologies" boilerplate changes the physics of meat, cheese, and bread.
When the market stops rewarding the performance, the mentions will vanish from the filings as quickly as they appeared. The risk factors will be quietly rewritten. Danny DeVito will go back to yelling about "the juice" without a single reference to neural networks.
Until then, every "AI" mention in a sandwich shop prospectus is a flashing neon sign: the hype has eaten the reality. Caveat emptor.