How Waymo, OnlyFans, and a 27-Year-Old HVAC Business Explain the Future

Nothing has just one consequence.

That’s the idea behind 1st, 2nd, and 3rd-order thinking… Thinking about consequences.

French economist Frédéric Bastiat was writing about this almost 200 years ago.

“The first [effect]… is immediate,” he says. “It appears simultaneously with its cause; it is seen. The other effects emerge only subsequently; they are not seen; we are fortunate if we foresee them.”

Nearly 2 centuries later, billionaire investor Ray Dalio calls this one of the most important principles of success in life:

1st and 2nd-order consequences often have “opposite desirabilities,” he says, “resulting in big mistakes in decision making.”

Or as Bastiat puts it: “It almost always happens that when the immediate consequence is favorable, the later consequences are disastrous, and vice versa.”

Like going to the gym. Pain first, abs second.

Or the 11 pm drive-thru run. First-order: transcendent. Second-order: why the hell did I do that?

Dalio and Bastiat would tell you that nature usually penalizes the people who stop at that first order.

Especially if you're building anything right now, first-order thinking won't cut it.

And honestly, the second and third-order stuff is way more fun to think about anyway.

Here's what we mean.

In 2026, Microsoft, Meta, Alphabet, and Amazon alone will spend ungodly amounts of money on AI infrastructure. Like, around $75,000,000 every hour.

And historically, when you do that, demand doesn’t hold steady. It explodes.

19th-century economist William Stanley Jevons noticed that when coal technology got more efficient, total coal consumption went up, not down. More efficiency created more demand.

Often it creates more jobs, too, in unexpected ways.

As Box CEO Aaron Levie put it:

“If you described Figma or Google Adwords to someone in the 1970s, they'd have predicted marketing jobs would collapse. One person could suddenly do what used to take a whole team. But there were a few hundred thousand marketing-related jobs in the US in the 1970s. Today? Low millions…”

How?

“Because advertising went from being something only your CPGs and car companies could afford to something any small business could participate in… More efficiency didn't kill demand. It created a massive new class of participants.”

We expect similar things to happen with AI.

Everyone’s heard the take by now: trades are AI-proof.

Plumbers, electricians, HVAC techs. Safe forever.

And we agree - to a point. AI won’t “replace plumbers” anytime soon. But don’t be so sure about “forever.” Even the trades will look different in the future…

As for the near-term second-order effects of AI, investor Michael Burry (AKA, this guy from The Big Short…) recently made an interesting point.

A middle-class homeowner facing an $800 plumber call, he says, might now prefer to snap a photo, ask an AI what to do, and try handling it themselves.

Obviously, that only works for the simplest stuff. But at scale, it could create a K-shaped split:

The top end of the trades pulls away from the bottom.

Again, just a second-order theory.

But it’s the reason why highly profitable “Lindy businesses” are having such a moment right now. The longer something has been around, the longer it is likely to be around.

And it’s why memes like this go so viral:

Because it’s true.

A 27-year-old HVAC company in Omaha with no website isn't sexy. But if your furnace goes out next week, AI isn’t going to replace it. Technicians will.

Remember our friend Frédéric? Here’s another banger he wrote:

Self-driving cars might be the purest example of this playing out in real time.

The near-term evil: Uber drivers losing their jobs. That's the first-order effect everyone fixates on, and it is unfortunately going to be painful.

But have you thought about the good to come?

Consider this. Through September 2025, Waymo cars have driven 127 million rider-only miles without a human driver. Here’s the safety record:

The second-order effect? If every vehicle in the US drove like this, we would prevent tens of thousands of deaths every year and likely save trillions of dollars across society. That includes our healthcare systems, which see around 3 million emergency visits every year for motor vehicle crashes.

A third-order effect? Auto insurance completely changes.

That’s already starting. Lemonade Insurance just announced a 50% rate cut for Tesla vehicles when Full Self-Driving is active. They even make use of second-order thinking in their marketing. Take a look:

Their co-founder put it plainly: “A car that sees 360 degrees, never gets drowsy, and reacts in milliseconds can’t be compared to a human.”

One technology. Hundreds of sectors reshuffled.

Nikita Bier (the Head of Product at X) made a prediction this week:

If that happens, what becomes more valuable?

Cybersecurity. Real relationships. Warm intros. Handshakes. Verifiable trust.

One recent survey found 58% of professional photographers in the UK had lost an assignment to AI by early 2025. Headshots, product shots, stock images, real estate photos. All are getting disrupted.

But weddings, news, sports, nature, and high-end photography? All could become more valuable precisely because of the human involvement and proof of reality.

The one nobody wants to talk about.

Loneliness is quietly becoming one of the most potent economic forces in the country. In 2023, Americans aged 15 to 29 spent 45% more time alone than they did in 2010.

In the coming years, businesses that create genuine human connections are going to have a massive tailwind behind them… as are those that capitalize on people’s loneliness in the wrong ways.

So what actually holds up in a world where the landscape keeps shifting?

One thing is EQ. Since 1980, people with high social skills have seen steady gains in both employment and wages, regardless of their math ability.

People with low social skills? Flat or declining, even if their math skills are strong. That trend was already accelerating before AI. It's probably about to go into overdrive.

The other thing is adaptability.

AI models that exist today will be obsolete in a year. Get comfortable learning new tools fast.

And that’s it. These things, along with ownership in a business, are the closest things to a durable advantage that exists right now. Practically, this means:

And now we want to hear from you:

What second-order effects are you seeing that nobody is talking about yet? What industries are about to get reshuffled in ways people aren't expecting? Reply and let us know.

Some of the best ideas we write about come straight from you.

-Team Contrarian