AI Isn't Taking Dealer Jobs. It's Ending the Ones That Shouldn't Have Existed.

There is a moment in 1991 that explains almost everything happening in automotive retail right now.

A family sitting around a desk in a home office. A slow dial-up connection screeching its way to the internet. A website loading pixel by pixel. And a father, turning around in his chair to look his kids in the eyes and announcing, plain as anything: this is going to kill the encyclopedia salesperson.

He was right. He was also, though he did not say it out loud yet, holding the death warrant for his own business. The family published phone books for a living.

That father was Michael Cirillo's dad. And the reason Cirillo tells that story now, decades later, as a dealer marketer who has watched AI eat into the foundations of everything he built, is that he has been here before. The feeling of watching a technology arrive and knowing it is going to change everything, while having no idea yet what to do about it. The defensive posture. The denial. The slow, grinding acceptance. The pivot.

He went through the whole sequence again when digital ate print. He is watching the whole industry go through it right now with AI. And his argument is that most people are stuck at the wrong end of it.

Two failure modes. Both will cost you.

There are two kinds of people showing up to every AI conversation in automotive right now, and Cirillo names them directly. The hype crowd and the zealots. The hype crowd tells you that AI is about to hand you a life of leisure, that the agents will do everything while you go to the gym and live your best life. The zealots tell you millions of jobs are already gone and there is no floor on how far it goes.

Both are narratives. Both are incomplete. And both are keeping people from doing the one thing that actually matters, which is using the technology in front of them.

The hype bothers him in a specific way. There is a Claude ad he breaks down on this episode, a slick little spot in which a busy entrepreneur rattles off a list of complex tasks to his AI on the way out the door, builds a website, pulls Instagram analytics, writes a strategy report, organizes the desktop, and then comes home and sits down to review. The capability is real. The direction the ad points is a little off. Because here is the thing the ad glosses over: the guy is clearly fluent in what he wants. He knows what a pixel-perfect website requires. He knows what analytics to pull. He knows what a strategy report should contain. He is not a passenger. He is a director. The AI is only as useful as the human pointing it.

That is the whole argument, and it is the thing the hype narrative consistently buries.

"AI scales the human. The question is what it's going to speed up."

What it actually scales

If AI scales the human, and the human is operating at a low level, AI scales that too. This is not a comfortable thought for an industry that has historically been fine with mediocrity on the floor, in the BDC, and in the tower. But it is the honest version of what is coming.

The flip side is the one worth focusing on. For the dealer principal who has built real operational knowledge, or the sales manager who understands the through-line of why customers actually buy, or the GM who has learned to read a store's culture with precision, AI is not a threat. It is a multiplier. It takes the ten hours of slogging through OEM scorecards and email threads and puts a ten-minute brief on the desk instead. And that time does not disappear. It gets redirected to the one thing every operator says they want more of and almost never gets: actually being with their people.

John Osborne, a VP at Carter Myers Automotive, has already built this. He connects Claude to his email and calendar, points it at the OEM attachments and scorecards that arrive every day, and has it synthesize a morning brief. What used to be, conservatively, a half-day of work done well, is now a ten-minute read. He walks onto the floor as an architect, not a processor. He pours his time and energy into his people instead of into documents.

That is not a hypothetical. That is Tuesday.

Three things you can do before the end of the week

Cirillo is specific about where to start, because specificity is what separates the people who actually move from the ones who just stay worried.

The first move is the morning brief. Connect your AI model to your email and calendar. Point it at the OEM reports and scorecards piling up in your inbox. Ask it to summarize, flag what matters, and surface the things that need your actual judgment. Ten minutes of review instead of three hours of excavation, and the rest of the morning goes to your team.

The second move is for anyone selling cars, from the floor to the desk. Pull your last ninety to one hundred and eighty days of sold data from the CRM as a report. Drop it into Claude or whichever model you use. Then start asking it questions. What are the common vehicles? What are the common objections and how were they handled? What patterns show up in who is buying and why? This is the kind of analysis that, done by hand, takes days and usually does not happen at all. Done with AI, it takes an afternoon, and it gives you a brief that tells you more about your own customer than most salespeople accumulate in years.

The third is the one Charlie Spradlin talked about on a recent episode, using AI as a thinking partner for your own leadership development. Spradlin is mid-transition from operator to architect at Art Moehn Auto Group. He asked his model to interview him, to learn how he works, what he wants to build, and what is standing in the way. It surfaced the one thing he already knew but had not fully accepted. He is still doing everything himself. He needed to let go. The machine named it. That started the conversation. It is an unusual use, but an honest one, and Cirillo's point is that most people would not even think to ask their AI a question like that because they are still using it to find a restaurant.

The playing field is flat, right now

Here is the thing about this moment that Cirillo wants the defensive crowd to hear, and it is the most useful reframe in the conversation.

Every previous technology wave arrived with a head start baked in. The dealers who were already digital in 2005 had a lead over the ones who were still all-in on print. The ones who figured out SEO in 2010 had a lead over everyone who ignored it. But AI arrived differently. It arrived all at once, in front of everyone, and almost nobody knows what they are doing with it yet.

"This is a once-in-thirty-years gift," Cirillo says, "where the playing field is completely level and all of us at the same time don't know what we don't know."

The window where that is true does not stay open. The dealers who start the unglamorous work now, the morning brief, the CRM analysis, the coaching conversations, the small daily use that compounds into fluency, those are the ones who will have a real edge when the window closes. The ones who waited for certainty, or for the hype to die down, or for someone to hand them a playbook, will have spent the window watching it close.

Cirillo's dad did not wait for certainty. He saw the encyclopedia salesperson's job disappearing, sat with that knowledge for a while, and then got out of phone books and into automotive magazines before the wave arrived. A few years after that, when digital came for magazines, he did it again. He built an inventory portal website that ranked in the top three in Canada. Not because he was fearless, but because he had learned, the first time, what it cost to be late.

The lesson was not about technology. It was about what you do when you see the writing on the wall.

Round it out

AI is not going to replace the car dealer. It is going to make the work of a car dealer much harder to hide from. The operators who are genuinely good at reading people, at building trust, at doing the real work of human connection, will find the technology clears away the busywork that was eating their best hours. The ones who were coasting on complexity, on being the only person who knew where the scorecards were, on the friction that hid their inactivity, will find the cover gone.

That is not a threat. It is a clarification.

The playing field is flat right now and the technology is sitting in front of everyone at the same moment. Start with ten minutes and a morning brief. Drop ninety days of CRM data into a model and see what it tells you about your own customers. Ask it a hard question about where you are stuck.

You do not have to be a technologist to start. You have to be willing to move before the window closes.

The encyclopedia salesperson did not.