Trading and AI: Efficiency, Danger, and the Pattern Trap

Trading and AI is not a simple topic.

It looks simple on the surface because the marketing is simple. “Smarter models.” “Better predictions.” “More edge.” But if you want to protect yourself, you have to know what you’re protecting yourself from. Most people are not afraid of AI itself. They are afraid of what change does to their sense of control.

Over the years, I’ve educated myself on this subject enough to have an opinion. Not because my opinion matters, but because in markets, naïveté is expensive.

And this is where it gets uncomfortable.

Fundamentally, AI is absolute madness because although we already know there is a 28% probability that AI will kill us all, we keep developing in that direction anyway.

At the same time, without AI it already barely works anymore. AI saves time. It compresses work. It scales output. It turns days into minutes.

Both statements can be true.

A few facts, without drama

AI is efficient. Undeniably.

AI is extremely dangerous.

A financial crisis, one that is certain to come, has very little to do with AI.

And yet, AI is fundamentally wrong. Why?

What AI actually does

So far, AI can only do one thing extremely well.

Pattern recognition.

It is like Google, but not on one track. On many tracks at once. It recognizes structure in chaos, correlations in noise, and it does so at a speed and scale that the human brain cannot compete with.

That’s the attraction.

That’s also the trap.

The deeper problem: humans are wrong about patterns

Human beings are generally wrong to limit, or rely, their causal system on pattern recognition.

There is a principle here that matters more than any model.

Reality is not a loop. Reality expands. Every moment is new. (If you want a metaphor, think Big Bang theory. Expansion, not repetition.)

But because the brain cannot process the raw volume of reality, it cheats. It compresses.

It says:

• “I already know this. It’s the same as three days ago.”

• “This new person looks like Fritz.”

• “This steak tastes like the steak from last week.”

And factually, that is wrong.

Every moment and every experience is absolutely new.

So pattern recognition carries much less weight than people think. It is often useful. It is often adaptive. But it is not truth. It is compression.

And here’s the uncomfortable implication.

If AI is a pattern engine, and humans are pattern engines, then AI is amplifying the same underlying limitation that already distorts most human thinking.

AI captures only a fraction of what is real.

Humans capture only a fraction of what is real.

And the mass of people, including the programmers of AI, usually believes that the fraction is the whole.

That is why this new religion will eventually collapse rather miserably. Not because AI won’t work, but because it will work inside an illusion people confuse for reality.

What this means for trading

Markets punish illusion faster than most areas of life.

Pattern recognition is everywhere in trading: indicators, setups, backtests, “this looks like 2008,” “this is the same as last cycle.” Some of it helps. Much of it creates false confidence.

AI will make this temptation stronger.

Because it can manufacture convincing “certainty” at scale:

• more signals

• more correlations

• more narratives that sound scientific

• more backtests that look clean until they fail live

AI will not remove human bias. It will industrialize it.

The edge is not having more patterns.

The edge is knowing the limits of patterns.

Why I’m less interested in opinions now

At some point I lost interest in opinions, politics, and a lot of the noise from the outside.

Not because I’m above it. I’m not.

But because I recognized something simple.

A lot is being lied about.

And a lot is rooted in misunderstanding.

Everyone invents something, believes they are smart, and babbles on while always chasing money or power. I’m certainly not smarter. But it saves time if you build your Lego house from principles.

Principles are the definition of absolute truths.

If you build your house of cards from absolute truths, you can rely on it. At least you have a foundation.

Everything is unstable. But houses of cards built from principles don’t have to rely on hearsay. They are, fundamentally, at least a bit more stable.

The real efficiency play

Proceeding this way first creates a small extra effort.

Instead of relying on newspapers or the loudest information source, you do the slower thing: you become interested in truth.

But over time, it becomes more effective. More time-saving. More protective.

Because you are no longer outsourcing your worldview to a system that is optimized for attention, persuasion, and patterns.

And that is the core point:

AI is a pattern amplifier.

If your life, your trading, and your decision-making are built on patterns without principles, you will be efficient at being wrong.

If your life is built on principles, AI becomes what it should have always been:

A tool.

Not a truth.