AI isn't as much of a bubble as you think..

Ever wondered why everyone is calling AI a bubble, but investors continue to invest astronomical amounts? Here’s one of the key reasons why, that you may not know.

If you look at the stock market or your LinkedIn feed, it’s easy to think we’re in a massive AI bubble. We’ve seen this movie before (the dot-com crash, anyone?). But if you look under the hood of how these models actually work, you’ll find a structural shift in how humans use computers - one that actually solves a looming crisis in physics.

The Atomic Wall

For decades, we’ve relied on Moore’s Law: the idea that we can just keep cramming more transistors onto a chip to make computers faster. But we’ve hit a wall. Transistors are now approaching the size of a single atom. When things get that small, electrons start jumping across gaps where they shouldn't (a phenomenon called quantum tunneling).

In short: we can’t just "brute force" our way to better performance anymore.

The Old Way: Raw Computation

Traditionally, if you wanted a computer to render a realistic 3D image of a sunset over the ocean, it had to do an absurd amount of math. It would calculate the path of every individual ray of light, how it bounces off the water, and how it refracts through the atmosphere.

This is Raw Computation. It is precise, but it is "expensive" in terms of time and electricity. It’s like trying to navigate a city by measuring every single crack in the sidewalk to find your way home.

The AI Way: The Power of "Intuition"

This is where the recent breakthrough in Transformer models changes the game. AI doesn't work like a calculator; it works more like a pattern recognition engine.

Think of it this way: If I ask you to draw a picture of a cat, you don't calculate the physics of how light hits fur. You don't solve any equations. Instead:

1. You have a "mental model" of a cat based on the thousands of cats you’ve seen in your life.

2. You use intuition to place lines and colors where your brain knows they "belong."

This is exactly what modern AI does. By training on massive datasets, the AI builds a statistical "map" of reality. When you ask it to generate an image or write code, it isn't "calculating" the answer from scratch; it is predicting the most logical result based on its internal map.

Why This Isn’t a Bubble

The "real" reason AI is here to stay is efficiency. Because AI uses "learned intuition" rather than raw math, it can achieve outcomes that were previously too computationally expensive to even attempt. We are moving from an era where we tell computers how to calculate a result, to an era where we show them what the result should look like.

The Takeaway: We haven't just built a faster computer; we've discovered a way to bypass the physical limits of hardware by changing the way we solve problems.

The limits of the "Atomic Wall" have been lifted. We’re no longer just crunching numbers; we’re teaching machines to understand the patterns of our world. That isn't a bubble - it’s an evolution.

Check out how Glow uses AI to empower humans, rather than replace them.

Not sure what Glow is? Think n8n, but for non-technical users.