Print the page
Increase font size
AI, the Very Different Beast

AI, the Very Different Beast

Chris Campbell

Posted August 21, 2024

Chris Campbell

People love comparing the AI boom to the Internet boom.

It’s like a reflex.

They’ll pull out graphs, throw up timelines, and slap on a big “YOU ARE HERE” label as if we’ve been here before.

The problem? We’re not just following an old map; we’re navigating uncharted territory - and with a GPS that keeps recalculating.

The Internet is a network. It’s a way to connect computers and people. It doesn’t do anything beyond that.

AI? It’s something else entirely.

Internet analogies will only take us so far.

Allow me to explain - and reveal why it has EVERYTHING to do with HOW you invest in AI.

Rise of Empires

Back in the early days of the internet, there was a ton of confusion about what it was.

The Internet wasn’t just a new gadget—it was a network. And networks are all about connections, feedback loops, and the endless chase for network effects.

The internet revolutionized the world because it made connections possible.

You had computers that could talk to each other, and eventually, people got really good at building applications on top of these networks.

The more people joined, the more valuable the network became.

Think about it: every additional user meant more content, more data, more people to connect with. That’s how platforms like Facebook, Google, and Amazon took off—they leveraged these network effects to build empires.

But AI?

AI is a Different Beast

If the internet was all about networks, AI is like the brain behind the network—a computer, but a very different kind of computer.

Traditional computers are deterministic: they do what you tell them to do, over and over, without fail.

They’re the reliable workhorses of the tech world.

But AI?

It’s probabilistic, which means it doesn’t always give you the same answer twice, and it might even argue with you.

In short, AI “thinks”.

It does so by processing data in a way that mimics certain aspects of human thought, but in a more mechanical and mathematical way.

In a way, this makes AI more like the original mainframe computers than the Internet.

Think about it.

Both AI and mainframe computers are specialized technologies.

The big mainframes were massive, expensive, and only a few organizations could afford them. The idea back then was that the world only needed a handful of them, and that was that.

That’s roughly where we are with AI.

Big players are competing to have the biggest, strongest, fastest, best.

But technology evolved, and computers shrunk in size, cost, and accessibility.

From mainframes to mini-computers, to personal computers, to smartphones, and now to everything from thermostats to cars—they’re everywhere.

Everything has a chip now. That’s where AI is headed.

The Everything Brain

Mainframe computers once dominated by being large and powerful.

AI is currently following the "big is best" model.

But, as with computers, the future of AI is likely to shift toward smaller, more specialized systems that integrate into everyday life.

These "brains" will be embedded everywhere, from the massive models of today to the countless smaller, more efficient ones of tomorrow.

If there’s any analogy to the Internet, it’s this:

You won’t just be using one big AI model—you’ll be using a network of them, all suited to your specific needs, running at different scales and with different capabilities.

And, AI isn't entirely devoid of the Internet’s dynamics.

As AI models are trained on more data, they improve in accuracy, creating a feedback loop similar to what we saw with the Internet.

So, while AI is more about intelligence than connections, it still benefits from network effects in its development and application.

Here’s the biggest difference:

Rise of the Super Expert

Recap: With the internet, the primary societal shift was about connectivity.

Bringing people, information, and businesses together in ways that flattened communication and knowledge barriers.

This led to global social networks, e-commerce giants, and the information age.

The Internet was, by and large, a despecialization technology. These days, everyone’s an “expert” about everything. As a result, we have more “influencer”-style experts and fewer real experts.

AI is a “respecialization” technology.

True expertise can thrive through specialized models.

Not only will it be harder for fake experts to present themselves as authorities, as their knowledge can be tested…

It’ll give rise to the “super expert”: experts who can leverage AI to achieve deeper insights, faster decision-making, and more accurate results.

And that’s something new altogether.

How to invest? More on that tomorrow.

The Outsider’s Guide to 10x

Posted September 15, 2025

By Chris Campbell

This was back when $2 million meant you could buy a mansion in Manhattan and still have enough left over for a butler to press your socks.

How to Become a Pawn Star

Posted September 12, 2025

By James Altucher

I talked to Rick Harrison of Pawn Stars. Here’s why pawn stores are the best businesses ever.

3 Boring Reasons Tokenization Goes to TRILLIONS

Posted September 11, 2025

By Chris Campbell

Don’t think of tokenization as some abstract blockchain thing. Think of it as moving from three buttons to an entire spaceship dashboard.

Nasdaq Tells SEC: “Tokenize Stocks Now!”

Posted September 10, 2025

By Chris Campbell

Six years ago, Nasdaq thought tokenization would start at the edges. Instead, it’s rapidly remaking the core.

Brussels to Beijing: Adapt or Be Dollarized

Posted September 09, 2025

By Chris Campbell

The world’s money is being dragged, kicking and screaming, onto crypto rails. And nobody (NOBODY) is ready for what comes next.

Live Long and Ask Siri

Posted September 08, 2025

By Chris Campbell

When Apple spends big, they buy brand power, irreplaceable IP, or deep infrastructure. This tiny stock has them all. And here’s the most beautiful part…