August 11, 2026 | Read online

Don't Fund the Bubble.

Most of what's happening is garbage. The infrastructure underneath it isn't.

Welcome back to the Launch Key 🚀

What goes around comes around. I mentioned the SaaS-pocalypse this Spring as panic hit all public SaaS companies. A little time - like with most tech hype cycles - gives better perspective. Sure, some of the no-code names have been impacted as AI agents logically take over those duties.

But the deathknell for Adobe, Salesforce, Oracle, SAP, Microsoft and others was misguided. Those companies are infrastructure most likely to not be vibe coded away.

Know which is which.

Let's get into it.

Read online

Gmail users may wish to read online since some parts may be clipped.
Take this week’s poll or comment below and let me know if we’re on the right track.

Pull to Eject
Modern Tools
Free Knowledge
Recommendations
Visual Crapshoot

Pull to Eject

I wrote a blog post in December 1999 saying the whole thing was nonsense.

Not because I was clairvoyant. Because the questions were obvious and nobody was asking them. I wrote that it wasn't the New Economy or the Old Economy — it was just the economy. Supply and demand. That "best to market" beats "first to market." That the whole dot-com hype rested on a dollars-for-eyeballs falsity, the belief that if you bought enough market share, profitability would obviously follow.

And I asked, at the time: how many pets do you know who shop for food on the Internet?

Within eighteen months, Pets.com was a Super Bowl ad followed by a liquidation sale.

Here's the part I got wrong, and it's the plot twist.

Many of those ideas weren't wrong. They were just early. Online pet supplies is now Chewy, a multi-billion-dollar business. Grocery delivery killed Webvan and then came back as Instacart. Same ideas. Same customers. Fifteen years apart.

So what changed?

The bubble paid for the road

Barry Ritholtz made a point on My First Million recently that reframed this for me: bubbles are good for the economy. Not for the people in them — for the economy.

$8B manager exposes the fake financial "Gurus" destroying your net worth

Speculative manias finance infrastructure that a rational market would never bankroll. The speculators get wiped out. The infrastructure stays behind, cheap and abundant, for whoever is still standing.

British railway mania in the 1840s ruined a generation of investors. The track they laid is still the backbone of British rail.

The dot-com bust did the same thing with bandwidth. Global Crossing laid enormous quantities of undersea fiber and went bankrupt doing it. That "dark fiber" is what carries your Netflix and this email. Bandwidth costs collapsed in the years after the crash.

And that is why grocery delivery eventually worked. Webvan wasn't a bad idea badly run. It was a good idea running on infrastructure that hadn't been paid for yet. The bubble paid for it. Instacart drove on the road Webvan's investors bought.

Now it's AI data centers. Hundreds of billions pouring into compute warehouses and the power to run them, on a bet about a future nobody can describe precisely.

History says most of that capital doesn't come back. History also says the concrete, the chips, and the transmission lines don't disappear when the money does.

Which one do they mean?

Here's the distinction almost nobody is making - the road is not the traffic.

The infrastructure layer — compute, power, the frontier models themselves — is expensive, overbuilt, and probably durable. It will likely outlive the people currently funding it.

The stuff being built on top of it is still mostly slop.

Those are two completely different bets, and every "AI bubble" headline mashes them into one word. So when you read one, ask which they mean. If they mean the traffic — the four hundred thin wrappers, the AI features bolted onto products that didn't need them, the pitch decks with a chatbot stapled to the side — they're right, and it's going to get worse before it clears.

If they mean the road, history is not on their side.

The pushback you're seeing right now is real and it's earned. But it's aimed at the traffic. Don't let it talk you out of the road.

Four questions that still work

You already own the instrument for this. It's the same one that worked in 1999, and it doesn't require you to understand fabs.

Who actually pays — and would they pay if it weren't free? Nearly every AI product you've been shown is subsidized by someone's venture round. Usage isn't demand.

Is this best to market, or just first? Google wasn't the first search engine. It was the one that worked. First-to-market was the 1999 delusion and it's back, wearing new clothes.

Is the metric real, or a proxy? Back then it was eyeballs and page views. In 2026 it's tokens processed, seats deployed, "AI-assisted workflows." Ask what business result sits underneath. If nobody can name one, there isn't one.

Take the AI out. Does the product get worse? This is the 2026 addition and it's the fastest of the four. If the thing still works with the AI removed, the AI is a press release. If it genuinely breaks — you're looking at something real. Not all of my AI experiments are going to make the cut.

That's the whole filter. And it will sort more of what's coming at you than any amount of technical knowledge.

Don't fund the bubble. Inherit it.

The anxious version says you're late. You missed it. Everyone else is already building.

The Big Picture says something close to the opposite.

You do not want to be early money in a mania. Early money buys the road for everyone else. That's not a tragedy — it's how the road gets built — but it's a job, and it's already taken. Somebody else is doing it, with somebody else's capital, right now.

The position you want is the one you're already in: still standing, watching carefully, with the questions above and no urgent need to bet the retirement account on a company that has "AI" in the name and no customers.

The people who got rich off the railways mostly weren't the railway speculators. They were the businesses that shipped goods on cheap rail afterward.

The last line I wrote in that 1999 column was that common sense makes a most welcome comeback.

It always does. Usually right about now.

Now go launch something 🚀

❝

Features do not scale, outcomes come due, hype fades, and results compound. If you focus on solving real problems faster than anyone else, the technology will take care of itself.

Jonathan Kirsten

CrowdHealth - Healthcare Crowdfunding

Tired of health insurance policies that are confusing and expensive? Join CrowdHealth and finally break free from the insurance system. CrowdHealth is not health insurance. It's an innovative model to use and pay for health care without insurance middlemen and government influence.

Modern Tools

EDGAR Full-Text Search — free, no account, run by the SEC.

Here's the thing about the four questions above: a public company will say almost anything in a press release, because a press release carries no legal liability. Their SEC filings do.

EDGAR lets you search the actual text inside every filing since 2001 — 10-Ks, 10-Qs, 8-Ks, proxies, exhibits, all of it. So when a vendor tells you their AI product is transforming the industry, go read what they told their investors. Search the company, open the most recent 10-K, and read the risk factors. That's where the quiet version lives: whether AI revenue is actually broken out as a line item, or whether it's folded into "other" because there isn't enough of it to name.

It's slow, unglamorous, and written by lawyers. That's the point. Nobody is hyping anything at you inside a 10-K.

SEC.gov | EDGAR Full Text Search

www.sec.gov/edgar/search

Free Knowledge

Ritholtz.com — Barry Ritholtz's blog, running since 2003, free, no signup.

The source for this week's issue. Ritholtz has been writing plainly about market psychology and why smart people lose money for two decades, and his central preoccupation is exactly our theme: the difference between what's actually happening and what everyone is saying is happening.

Start with his "15 Minutes" post from June, which links the My First Million appearance and the bubbles segment.

Recommendations

📕 MGMT Playbook : Practical management insights straight to your inbox every Wednesday.

🏢 Retirepreneur : Your guide to turning decades of experience into profitable consulting, coaching, or course businesses, without risking your retirement security.

🗃️ Dealroom Business Success Uncovered : Learn directly from billionaire entrepreneurs on how to grow a business. Join a community of 2,000+ innovators.

🧾 Redefining Retirement : Too Young to Retire, Too Old To Take Orders? Yeah, Me Too.

Visual Crapshoot

You've Spent 30 Years Building Expertise. It's Time to Turn It Into Income That Works for You.

The complete framework for professionals 50+ who want to build multiple revenue streams without quitting their job or betting everything on a startup.

Help me help you: What's the #1 challenge holding back your late career side hustle?

I'm creating resources specifically for late-career entrepreneurs. Which topic would help you most?

Share Launch Key | The Newsletter for Late-Career Entrepreneurs

Thanks for referring the Launch Key newsletter! 🚀

You currently have 0 referrals, only 1 away from receiving Free tools to launch your new idea or digital product. Instant PDF download with 1 referral.