• by Jordan Brannon , Yesterday

Vintage sci-fi movies feature alien crafts landing on earth.  Once the doors open, viewers find metallic interpretations of existing technologies.

That’s how marketers approach new advertising channels. We see a shiny object and expect something revolutionary.

What we get borrows the business model, language, and measurement habits of the medium that came before. It's not until people change their behavior that advertising really changes.

AI advertising will follow the same pattern.

In the near term, it will look a lot like search advertising because today’s AI interface still looks a lot like search. A user types into a box, the system responds, and adds a layer in around the experience.

AI advertising becomes meaningfully different only when it knows your calendar, location, apps, purchases, routines, photos, inbox, and preferences.

The First Web Ads Borrowed From Print
The early web dressed up like old media so advertisers could understand it.

Banner ads were display ads with a modem. A rectangle with a headline, maybe a logo ending with a click. The technology was new, but the creative grammar came from print. AI advertising is beginning the same way with sponsored answers, promoted links, and shopping recommendations. Google has expanded Search and Shopping ads into AI Overviews [1]

OpenAI’s ChatGPT presents product recommendations with images, and reviews.

New media begins with imitation. The early web borrowed from print. Early AI ads will borrow from search.

Search Proved  Declared Intent
Search turned declared intent into money.

A person typed “best running shoes for flat feet” and the system didn’t need to guess commercial intent. The user declared it. The advertiser was no longer buying a vague audience segment. They were buying a moment.

 Prompt-based advertising extends that model. The query gets longer, and more conversational: “I run three times a week, I overpronate, I want something under $150, and I hate bulky shoes.”

Google’s AI Max for Search captures complex, long-tail searches that older keyword structures  miss. The prompt is  the new keyword. [2]

AI advertising will feel like a rerun: the system matches intent to an advertiser, ad displays,  action is tracked, and advertiser pays.

Free AI tools add pressure to that model. AI answers cost more than serving links. Heavier use means stronger incentive to monetize intent. Search proved a free utility can become a highly profitable ad model. AI assistants are lining up behind the same toll booth.

The Smartphone Changed Advertising By Changing The Interface
The smartphone changed advertising because it changed behavior.

Desktop search meant sitting down, typing, browsing, comparing.

The smartphone added location, maps, cameras, payments, and immediacy. “Hotel Tonight” became a bidding event.

AI needs its smartphone moment.

Today’s AI interface is still a text box with better manners. Useful? Yes. Revolutionary in advertising? Not yet.

AI ads will feel truly new when the interface changes.

Voice, wearables, cameras, agents, cars, glasses, and ambient assistants create different behavior. A user sharing contexts just by existing will make ads something else.

Google has made Search more exploratory and multimodal, through AI Overviews, Lens, and Circle to Search. [3]

Until that happens at mass scale, AI ads will mostly be search ads wearing a nicer jacket.

Social Created Different Signals
Social advertising became important because people created new signals

Likes, follows, shares,  group membership, and engagement  gave ad systems a different map of behavior. Search knew what someone wanted in a moment. Social knew what someone watched, joined , and performed for others.

AI will create its own category of signals as it grows beyond prompt context.

A personal AI could know the trip on the calendar, the birthday in contacts, the fitness goal in the app. That's a different ad signal.

This is where AI  becomes more powerful than search and more sensitive than social.

A user asks, “Book the best hotel for my trip,”" or "Pick a birthday gift .” The assistant isn’t presenting options. It’s narrowing the field, maybe completing the task. The question shifts from “Where does the ad appear?” to “Who decides?”

That's where trust becomes the product. A sponsored recommendation from a personal assistant that knows your budget, health and family is another beast.

The Rerun Comes Before The Rupture
AI advertising will repeat history in two ways.

First, it will borrow from old models before inventing new ones: sponsored answers, promoted recommendations, conversational keyword targeting. The prompt is the new keyword.

Second, the real winners will be companies that control the interface where behavior changes.  A search engine, AI assistant, operating system, device, or agent layer between the user and the internet. Whoever owns the user relationship, context, and measurement wins.

The longer-term future will look less like ads next to answers and more like advertising attached to personal context and task completion. Something closer to personal assistants who’ve been bribed.

Or perhaps something more alien than we can predict.

[1]: "New ways AI in Search helps your business"

[2]: "Introducing AI Max for Search campaigns"

[3]: "AI, personalization and the future of shopping"

If you’re interested in submitting content for future editions, please reach out to our Managing Editor, Barbie Romero at Barbie@MediaPost.com.