AI has to be experienced

This sponsored article by AI Digital explores why hands-on experience is essential to understanding AI’s potential and limitations.

Written by: David Mainiero, Chief AI Officer, AI Digital

Andrej Karpathy, an OpenAI co-founder and Tesla’s former head of AI, recently described the gap in how people understand AI as a funnel that keeps getting sharper. About three-quarters of the world has barely touched an LLM. Another billion or two use free chatbots now and then, and mostly see what he calls a “derpy chatbot,” a better Google search. A much smaller group, mostly engineers, is watching projects that used to take weeks get finished from a single prompt.

The line that stuck with me was about executives. They understand what’s happening intellectually, he wrote, but it’s still “second-hand and a bit abstract.” That describes most of the marketers and agency leaders I talk to. They’ve read the headlines and typed into a chat box. They haven’t had the moment where it clicks.

The numbers show it. Some 94% of mid-market companies use generative AI, but only 2% have operationalized it at scale, according to a May 2026 Kaufman Rossin study found. Agencies look similar: 61% of US agencies treat AI as a cost of doing business, and just 6% treat it as a line of business, according to a June 2026 study from Forrester and the 4A’s. People scope what they’ve touched, and for most people that’s a prompt and an answer.

That’s why on October 29, AI Digital is opening the AI Innovation Center in Miami. The Gensler-designed space isn’t a showroom. We built it to give people the in-person “epiphany moment” with AI that slides and webinars can’t.

Watching a demo keeps you a spectator. Using a system teaches you what it notices, what it assumes, and where it needs you to steer. That matters most for companies without a big in-house innovation lab, which is most of the market.

Most of the experiences center on creative work and future ad formats, and none involve typing into a chat box. Many feature the latest of our more than 40 AI Digital Labs incubator tools, and our flagship Elevate platform and its fleet of agents are built into some of them, so even the fun parts connect back to your actual work.

In one, a city rushes toward you, your body is the controller, and real campaigns run on the billboards you pass while a camera tracks your gaze. At the end it asks what you remember, and the answer is usually very little. Amplified Intelligence has found that 85% of digital ads never get the 2.5 seconds of active attention a memory requires. It’s one thing to read that stat and another to fail the test yourself.

In the creative studio, you bring the direction and AI turns an idea into finished work in minutes, with reactions from synthetic focus groups. When a new variant costs almost nothing, the hard part becomes deciding which one deserves the media budget.

In the campaign flight simulator, you feel cross-channel tradeoffs instead of reading about them.

Even the room is part of the lesson. It runs presentation tech you won’t find in the biggest companies’ briefing centers yet, and agencies can run branded takeovers of the space to bring their own clients through.

We’ll refresh the experiences on a rolling six-month cycle, because the technology moves too fast for anything else. The goal is to keep the AI Innovation Center at the state of the art, and to make that state of the art something you can actually feel.

There’s another reason to get hands-on. The questions that matter most about AI usually aren’t technical. What informed this recommendation? What’s missing? Where could it be wrong? You get better at asking them by using the thing.

A fair warning, though: using a system teaches you its strengths faster than its failure modes, and a slick experience can leave people more confident than they should be. So we try to show where things can break as clearly as where they shine.

The funnel Karpathy describes is getting steeper. In advertising, I think the agencies that pull ahead over the next two years will be the ones that move the most people down it: people who’ve used AI enough to know when to trust it and when to push back. That’s what the AI Innovation Center is for.

You've read 0 of 2 free articles this month.

Get more articles - create your free account today!