The Knife of AI Is Already at the Marketer's Throat

A few days ago, I saw a line in a marketing group chat.

It stopped me cold for a few seconds.

"Your job won't be taken by AI. It'll be taken by someone who knows how to use AI."

Brutal, right?

Brutal.

But it's true.

First, a Question for You

Have you ever wondered why that song Spotify pushes at you every day fits your taste so perfectly?

Why is the hit-rate on Netflix's "You might also like" row so absurdly high?

The answer is so obvious you'll kick yourself.

Behind the scenes, a team of marketers is using AI to watch you.

Watching when you tap, what you tap, how long you watch, the exact second you scroll away.

Then turning every one of those actions into fuel for the next push.

AI Isn't Coming — It's Already Here

When friends bring up AI with me, the first question is always the same: "Run-zong (a respectful Chinese nickname for a senior marketer — 'Boss Run'), can AI really be used in marketing?"

It really can.

And it's not "almost here" — it's already sitting next to you.

Those tools you open every day — the ones for email, social scheduling, customer-relationship management — they've had AI baked in for a while. You just might not have noticed. Or you noticed, but never dared to click in.

I went and dug through an industry survey from last year, and one line really stung: a lot of marketers say they "can't live without AI anymore."

But here's the catch.

The same survey found that the overwhelming majority of marketers are severely under-using AI.

What does "under-using" mean?

It's like buying a Ferrari and driving it to the grocery store every day.

The horsepower's there. You're just not stepping on it.

From "Guessing" to "Computing"

Let me break this down for you.

How did marketers used to work?

Look at demographic data. Look at surveys. Look at last year's numbers.

And then?

Guess.

Based on experience, based on gut — guess what this batch of people might like, make a batch of creative assets, push it out, and wait for the data to come back.

That playbook looks painfully clumsy in the age of AI.

Today's algorithms can compute, in real time, what you'll probably want next second — at the very instant of every click, every pause, every add-to-cart-then-delete.

Think about it.

One side reacts after the fact. The other side predicts before the fact.

One side blasts the masses. The other side is individualized, at scale.

Are those two even the same thing?

Let's Do the Math

Let me give you two examples.

Netflix. Everything you watch, everything you search, everything you abandon halfway — it remembers all of it. Then uses that data to reverse-engineer your taste, tag you, and recommend. The more you watch, the better it knows you; the better it knows you, the less you can leave.

Amazon. What you've bought, what you've browsed, how old you are, where you live — all fed into the model. The recommendations come out so accurate you can't help but click.

That's personalization.

And not for one person — for hundreds of millions of people, simultaneously, every second, nonstop.

In the old days, a marketing team working themselves into the ground could personalize for how many people?

A few hundred? A few thousand?

Today it's one person plus one machine, doing the work that used to take a thousand.

My god.

But Here's the Real Problem

Having said all this, I owe you a fair point.

AI is great, but it's also raised the bar.

A while back I was talking with a teacher who's been in the industry teaching for years. She said something I chewed on for days:

"What worries me most isn't AI replacing veterans. It's how newcomers are supposed to break into the field."

What does that mean?

Think about it. A marketing rookie, fresh out of school — what did they used to do?

Write email copy. Do basic design. Crunch data. Schedule the social calendar.

These are exactly the jobs AI is best at — exactly the first ones to get replaced.

So here's the problem —

If the entry-level work is all done by machines, where do newcomers train? Where do they build their chops?

That's the real question. Not "will AI take my job" — but "is the next rung of the ladder in this industry still steady?"

And a Few Pits You Should Know About

The knife of AI is handy, but it cuts easy.

Pit one: data.

What do those AI companies use to train their models? Stuff from the internet. Images, text, video. The problem is — a lot of it was taken without your consent. Your photos, your writing — they may already be sitting in some model's belly.

Pit two: bias.

Models learn patterns from data. But what if the data itself is biased? Like certain groups getting systematically under-recommended, under-exposed. The model won't correct itself — it'll just amplify the bias, then amplify it again.

That teacher told me, "On bias, we still have a long, long way to go."

She wasn't exaggerating.

Pit three: trust.

A legacy sports magazine, because it secretly used AI to generate content and didn't tell its readers — when it blew up, the CEO got fired on the spot.

See, the problem isn't using AI. It's being dishonest.

Readers can accept you using machines to help. Readers cannot accept being lied to.

So, What Now?

After all this, you might be asking: so what should I actually do?

Three things.

One, get your hands dirty.

Don't just read the news. Don't just listen to other people hype it up. Open those tools one by one, play with them yourself. Breaking them costs you nothing. Only after you've played do you know what's genuinely useful and what's just hype.

Two, understand the principles.

You don't have to write code. But you do need to know, roughly, how AI eats data and spits out results. Otherwise you can't even tell whether it's biased, let alone where.

Three, hold the line.

Using AI is fine — tell your users. When you mess up, own it. Taking data is fine — do it compliantly.

Stripped down to one line —

The tools change. The craft doesn't. And the way you do right by people doesn't either.

Finally

My teacher said one more thing. It's stuck with me ever since.

"AI won't replace people who know the craft. But craftspeople who can use AI will replace craftspeople who can't."

Let that sink in.

Really let that sink in.

This fight isn't person versus machine. It's person versus person.

And the one who knows how to use the machine has already started.

What about you?