Let’s stop pretending: most of what gets labeled “AI” in this business wouldn’t pass a CAPTCHA test, let alone solve real problems. Somewhere between the overpromised dashboards, cookie-dependent “targeting,” and 200-slide martech stacks, the word intelligence lost all meaning.

So when I asked Richard Howe, CEO of Inuvo, how exactly he plans to save advertising from itself, he didn’t blink.

“We’re going to save it by asking a different question that has been historically asked,” he said. “The question that has been asked for probably the last 30 years and has formed the basis for the way audiences have discovered and targeted is: who is someone?”

That worked—barely—back when the web was young, data was cheap, and nobody cared what you did with their browser history. But according to Howe, that entire premise is now obsolete. It’s crumbling like a cookie in a browser privacy update.

“With cookies going away and other privacy legislation coming in, that’s becoming obsolete. In fact, it’s becoming to the point where you can’t actually do it at scale anymore.”

So what question should we be asking instead?

“Why is it they’re interested in things?”

In other words, intent over identity. Motivation instead of metadata. And it’s not just a clever philosophical pivot. For Howe and Inuvo, it’s a whole new architecture of marketing built on something that sounds like science fiction but is already working in the wild.

One of the best examples of this “intent-first” mindset came from a campaign Inuvo ran for Bose. They were marketing Sleepbuds—tiny earbuds designed to help people sleep through noise. Pretty straightforward use case. Or so you’d think.

Traditional adtech would’ve gone after the obvious: travelers, insomniacs, maybe some stressed-out parents.

But Inuvo’s AI—trained on reading “everything that has ever existed,” as Howe put it—took a much stranger, smarter path.

“One of the things the AI figured out was that there is a group of people, call it an audience segment, where they have short-nosed dogs,” Howe told me. He wasn’t talking about Labradors or Goldens. He meant pugs. Bulldogs. Frenchies. The snorters.

“The AI was smart enough to know it’s not any dog—it’s short-nosed dogs. And the reason why it made that connection is because people who sleep with these dogs at night have trouble sleeping. They snort. They make a lot of noise. They keep people up.”

At this point, I had to interrupt.

“Yeah,” I said. “We have a half pug, half chihuahua. It’s a little bit like Dr. Jekyll and Mr. Hyde.”

Howe laughed. “Exactly. So the AI figured out there’s an audience here. A very specific one. You’re not going to find that in a CRM. There’s no checkbox that says: ‘Sleeps with Darth Vader-sounding pug.’”

Then it got even weirder.

“The other audience the AI picked up was Northern geographies—places like Iceland, Norway, Alaska. Why? Because the sun stays up longer. People can’t sleep. Add in a snoring pug and you’ve got a sleep-deprived market just begging for relief.”

You can’t target that with third-party data. You can’t track it with cookies. And you definitely can’t buy it from an off-the-shelf data broker. You need a system that understands nuance—motivation—the why.

The cookie, in Howe’s words, is “on its way to being dead.” Not just phasing out—dead. And not in the fun Halloween way.

Why? Because programmatic advertising, as it currently functions, is completely dependent on matching user identity. That one tiny identifier—an ID stored on your browser—is the connective tissue holding together the entire surveillance economy.

“There’s not a decision that gets made—at least in open web ad targeting—that doesn’t start with asking, ‘Is there a cookie ID?’” Howe explained. “And if there isn’t, they just skip the impression. It doesn’t transact. It’s worthless to them.”

This isn’t a minor inconvenience. It’s an existential crisis for anyone still clinging to “lookalike audiences” like a security blanket. The ability to onboard data, personalize content, and deliver performance all hinges on a string of characters that most users now block by default.

“If you don’t have the ID, you can’t onboard the data,” Howe said. “The data is useless.”

So, I asked the obvious question: Do traditional ad tech executives know they’re selling snake oil? Or are they just blissfully ignorant?

Howe didn’t hesitate. “Yes.” Just yes.

But it’s not stupidity, he clarified. It’s inertia. They’ve been doing it the same way for so long that they literally can’t imagine another model.

“It’s a paradigm problem,” he said. “They’re stuck. Like how train tracks are still four and a half feet apart because that’s how Roman chariot ruts were spaced. People just keep building on top of that without questioning the logic. And now you’ve got these giant trains trying to fit into ruts made for chariots. That’s adtech today.”

So what makes Inuvo’s AI so different? Why isn’t this just another machine-learning startup with a buzzword bingo card?

Because it’s built on what Howe describes as “theory of mind.” Not just reacting. Not just remembering. Actually simulating thought—context, intent, causality.

He broke it down into three levels of AI:

“Our AI is in the third category,” he told me. “It doesn’t just know what you did—it tries to figure out why you did it.”

That’s how it can make bizarre-but-brilliant leaps, like connecting a noisy pug in Alaska to a Bose Sleepbud sale. And it does it without violating privacy, breaking laws, or relying on identity resolution hacks that will be obsolete in two quarters.

🚧 Under the Hood: How exactly does Inuvo’s AI “read the internet,” and what data powers its insights?

💣 The Identity Collapse: Why legacy platforms are scared—and what real intent-based AI means for their business models.

📉 Myth-Busting “AI Adtech”: Howe explains the juice machine grift, the blockchain comparison, and how to tell when vendors are just putting lipstick on a rules engine.

👉 Subscribe to ADOTAT+ to read it first.
Because if your current AI strategy still involves a spreadsheet, a data co-op, and a bunch of expired cookies… you’re already behind.