AI is transformative. I want to say that up front so nobody accuses me of being a luddite. What multiple companies are doing right now across medicine, logistics, scientific research, and software development is real, significant, and genuinely reshaping how the world works.

That is not what 90% of the ad industry is doing.

What 90% of the ad industry is doing is renaming its seven-year-old bid optimization model, slapping ".ai" on the domain, and posting a LinkedIn carousel about "the future of intelligent marketing." And a staggering number of you are nodding along like you just received prophecy. You didn't. You received a rebrand.

Quick vocabulary lesson. Artificial intelligence is the big umbrella: systems that simulate human cognition. Reasoning. Planning. Creativity. Actual thinking about novel problems.

Machine learning is a subset: algorithms that learn patterns from data to predict outcomes. Click probability. Conversion likelihood. Bid price.

All ML is AI. Not all AI is ML. And virtually everything in your programmatic stack? It's narrow, predictive ML. It has been for years. The math didn't change. The pitch deck did.

Your "AI-powered" bid optimizer? ML. Lookalike modeling? ML. Creative selection picking which product tile to show? ML. Frequency capping and budget pacing with some forecasting bolted on? ML wearing a turtleneck.

These are pattern-recognition and prediction problems. They are not autonomous agents "thinking" about your brand strategy. They're doing math on behavioral data, which is exactly what they were doing in 2019 when nobody bothered calling it AI because it wasn't sexy enough for the Series C narrative.

Half the companies in your stack now end in .ai. The vast majority haven't changed a single thing about their product. Not the models. Not the architecture. Not the capabilities. They changed the URL. Maybe they added a chatbot to the dashboard that summarizes a report you could read in ninety seconds.

This is AI-washing. Same game as greenwashing, except instead of pretending to save the rainforest they're pretending to have invented the future. Everyone says "AI-first." Nobody means it. Your CPMs stay the same.

Let's call it what it is. The Cult of AI has all the features of an actual cult.

Special language and in-group status. "AI-vangelists." "Doomers." "Accelerationists." If you can drop "agentic workflows" at dinner, you're in the priesthood. Never mind that you can't explain what a gradient is.

Blind devotion to prophets. People are deferring real business strategy, real money, real headcount decisions to whatever some AI booster with 200K LinkedIn followers said in a carousel. Not based on incrementality tests. Based on vibes and a TED talk.

Quasi-religious fervor. Watch any adtech CEO describe their "AI vision" at a conference and tell me that's not a sermon. That's not a product roadmap. That's theology.

And here's why it matters: cultish narratives distort real decisions. They justify reckless rollouts. They replace risk-benefit analysis with faith. The hype is actively hurting the real thing.

OK. Now I need you to pay very close attention because this is the part that makes me genuinely angry.

The adtech industry spent the last fifteen years doing one thing brilliantly: making everything so unnecessarily complex that you couldn't question it. That was the entire business model. Wrap commodity technology in jargon. Build impenetrable dashboards. Create a Byzantine chain of intermediaries. Make the whole thing so confusing that when a CMO asks "where did my money go?" the answer is essentially "it's complicated, you wouldn't understand, trust us."

That was the grift. And it worked for years.

Now here's what's happening. AI, real AI, the actual technology, is threatening to undo that entire racket. Because suddenly you can interrogate data yourself. You can ask a model to explain what a DSP actually did with your budget. You can have a tool audit the supply chain, read the log files, surface the arbitrage. AI has the potential to make adtech legible for the first time in its miserable history.

And these people are terrified of that.

So what are they doing? They are lying to you. Again. The same people who made the ecosystem deliberately opaque are now repackaging that opacity as "AI." They're not adopting AI to help you. They're adopting the language of AI to keep you confused. To keep you dependent. To keep you from asking the questions that AI would actually let you answer.

I want to be very clear about this. I'm not being rhetorical. I'm not being provocative for clicks. They. Are. Lying.

All those old dudes my age who spent the last decade profiting from complexity and are now suddenly selling you a shiny new "AI-powered" product? They are liars. They don't have AI. They have a new coat of paint on the same black box they've been hiding behind since 2012. They know that if you actually understood what was inside, you'd realize you've been overpaying for math that a first-year data science student could replicate.

They made it complex so you couldn't question them. Now they see that AI could make you question them. So they're doing the only thing they know how to do: lying harder, faster, and with better production values.

The new product isn't new. The AI isn't AI. The innovation is the same innovation it always was: finding a more sophisticated way to make sure you never look behind the curtain.

And then, because the audacity has no ceiling, these same people are now charging you thousands of dollars to learn their fake version of AI.

"AI Guilds." "AI Academies." "AI Certified Professional" programs. I looked at how many of these exist in and around adtech and I almost threw my laptop. What is wrong with all of you paying these people?

Here's the play: they're inflating the difficulty on purpose. "Mysterious, hard, and rare" AI is easier to monetize than "statistical tooling wired into workflows." If they told you the truth, that most of what matters for your daily work is learnable in a few weeks, you wouldn't pay $3,000 for a badge.

So they sell you fear. There's a catastrophic "skills gap." Without their certification, their proprietary framework, their LinkedIn badge, you will be left behind in the great AI rapture. It's a fear sale layered on top of a lie layered on top of a grift. It's grift lasagna.

What do you actually get? Vocabulary. Slideshows. Multiple-choice exams. You come out able to say "transformer architecture" at a meeting.

You cannot build or operationalize anything.

Then they hit you with renewal fees, "continuing education" dues, event tickets, and tool bundles.

One certificate becomes a subscription to your own insecurity.

The people selling these courses are, in many cases, the same people who spent the last decade making adtech deliberately incomprehensible. They are not educating you. They are doing what they have always done: monetizing your confusion. The subject just changed from "programmatic" to "AI."

Most of "the AI" that matters for day-to-day work is surprisingly easy to learn.

AI literacy (understanding what ML and LLMs do, where they fail, how to use the tools) takes a few hours to a couple of weeks. Free resources everywhere.

Practical power-user skills (wiring AI into workflows, prompting, automations) take a few weeks. You do not need a $4,000 bootcamp.

Actually engineering models from scratch? That takes 6-12+ months. That part is hard. But that is not what you need to do.

The public conversation deliberately conflates "can I use AI effectively?" with "can I build a neural network from scratch?" because the confusion is profitable. If AI feels like a PhD, you need their course. If it feels like learning new features in Excel, which for most use cases it basically is, you don't need them at all. Here’s a starting point

The adtech industry has a long, proud history of taking real innovations and turning them into meaningless buzzwords designed to justify fees. We did it with "programmatic." We did it with "blockchain." We did it with "the metaverse."

Now we're doing it with AI. The difference this time is that AI actually works. Which makes the bullshit even more unforgivable.

We have something genuinely powerful. Something that could bring transparency to an industry that has fought transparency at every turn. Something that could actually empower marketers to understand where their money goes. And the old guard, the same people who built the opacity machine, are hijacking it to build a newer, shinier opacity machine.

They made it complex so you wouldn't question. They see AI could make you question. So they're lying again.

I am tired of it. You should be tired of it too.

Stop buying the cult. Stop paying the liars. Start using the actual tool. It's not that hard to learn. It's not that mysterious. And it sure as hell isn't what they're selling you.