What Makes People Fall in Love with Someone Who Doesn’t Exist?

I’m building an AI influencer named BoxxBoxx. A blonde woman in a black racing suit who makes music for Formula 1 fans. She doesn’t exist, but I want people to care about her anyway.

That’s why I drove to West Friesland to pick up Patrick Kicken. He’s my first guest in this new podcast series. He’s also the only person who’s been a guest three times across all my different podcast series. There’s a reason I keep coming back to him.

Patrick understands what I’m struggling with. He spent decades as a radio DJ, creating connections with listeners through nothing but his voice. He’s seen musicians explode into stardom while others, equally talented, never break through. He knows what makes people tune in, follow along, and actually care.

I needed to ask him the big question: Can people feel the same emotions for something that doesn’t truly exist?

The Harry Potter Test

“Can it work?” I asked Patrick as we drove through the Dutch countryside. “Can people become fans of someone they know is fictional?”

I shared my thinking: “People are fans of Harry Potter. Of Homer Simpson. Of a lion named Simba.”

Patrick nodded and added more examples. “Peter Griffin from Family Guy. Mickey Mouse doesn’t exist either.”

That’s exactly my point. We fall in love with characters all the time. Children cry when Mufasa dies. Adults get emotional watching Elsa sing in Frozen. These are animated drawings, yet they move us.

So why wouldn’t it work with BoxxBoxx?

The Albert Heijn Lesson

Patrick dropped a marketing truth bomb early in our conversation.

“People respond to people, not to brands,” he said. “That’s why Albert Heijn was so smart with their supermarket manager character. That bald guy in all their commercials. He was just an actor, but he became the face of Albert Heijn.”

We saw his personality. His humor. His quirks. And we connected with that.

That’s what I need to do with BoxxBoxx. But here’s the challenge: I haven’t fully defined who she is yet.

I know she’s around thirty. She has a slight British accent. She’s a digital nomad, always traveling to Formula 1 circuits around the world. She wears that black racing suit everywhere.

But that’s where it ends. And Patrick immediately called me out on it.

The Cardboard Cutout at Veronica Radio

“At Veronica Radio,” Patrick told me, “we literally had a cardboard cutout in the studio. It showed our ideal listener. A man around thirty, living with his partner, starting to think about kids. Every piece of content had to be tested against that person.”

He paused. “You need to create that same kind of profile for BoxxBoxx. How old is she exactly? Where does she come from? What’s her backstory?”

I realized I’d been so focused on the technical execution that I’d skipped the foundation. The tools work. Suno creates incredible music. VEO3 from Google makes videos that are almost indistinguishable from reality. ElevenLabs can clone any voice.

But all that technology means nothing if the character isn’t real to people.

Show the Vulnerable Side

Patrick shared something crucial about building connection: “Let her be vulnerable sometimes. Let her cry. Show her struggling with something.”

He pointed to reality shows. “Look at Gordon showing off his luxury yacht in Dubai versus Gerard Joling’s soap opera. People connect with Joling because they see him getting frustrated with his manager, making eggs at home, being human. Not just the glossy side.”

This hit me hard. I’d created one humorous video for BoxxBoxx that completely flopped. Looking back, I’m not even sure it fit her character. Because I don’t have her character clearly defined yet.

Patrick was adamant: “Stay true to your character. The worst thing you can do is be tough one week and timid the next.”

He referenced how boy bands were constructed. “Backstreet Boys, Take That, NSYNC. The record labels literally drew it out on a board beforehand. One shy guy, one tough guy, one great dancer, one with curls. They designed the personalities before they even found the people.”

I need to do the same for BoxxBoxx.

The Music Has to Be Good

We listened to my latest track in the car. It’s about McLaren potentially winning the constructor’s championship. Patrick’s verdict was encouraging but came with a reality check.

“I think it’s smart,” he said. “You can respond to current events much faster with AI. But…”

He paused.

“…it has to start with the content. If you come with mediocre songs and try to make the character popular, it won’t work. The music itself has to be good.”

That’s the foundation. Everything else builds on top of it.

The X-Factor Question

We kept coming back to this mysterious thing. That intangible quality that makes certain people magnetic.

“It’s a feeling,” Patrick said. “Humor first. Edwin Evers became successful because he was like the ideal son-in-law. Friendly, someone you’d want as a friend. He creates the right atmosphere.”

For a woman targeting a male audience in Formula 1, Patrick suggested focusing on relatability. “Let her show what kind of car she drives. Show that she loves driving fast. Build things around her that the audience connects with.”

But there’s more to it than just strategy. Patrick mentioned voice quality, timing, consistent personality, sharing personal details over time. And that thing that’s hardest to define: likability.

The Collaboration Strategy

Patrick emphasized this multiple times: “Collaborations. That’s how Milan Knol became famous. Through his brother Enzo Knol. The Roelvinkjes all helped each other grow.”

This is already my number one approach. I’m planning a podcast episode with Rudy van Buren, who works at Red Bull Racing as Max Verstappen’s simulator driver. He’s one of the best sim racers in the world and is well-connected in the F1 paddock.

I’m also in talks with Alexander from Grand Prix Radio. If they occasionally played BoxxBoxx’s music, that could be the breakthrough moment.

Timing matters too. I released the McLaren championship song at the start of race weekend, hoping it would catch the moment when they actually win. We’ll see if it works.

Can AI Replace the Real Thing?

Our conversation naturally drifted toward the bigger implications. Patrick shared that radio stations in the Netherlands are already using AI newsreaders on weekends.

“I think in ten years, daytime programming will mostly be AI voices,” he predicted. “Only the morning and afternoon shows will have real humans.”

He’s surprisingly comfortable with this. “I’ve had my voice cloned. I got paid for it. If it modernizes the industry, why not?”

But I pushed back on one point. For voice actors who do this full-time, losing that income stream is real. Patrick acknowledged it but compared it to historical precedent.

“When records were invented, orchestras protested. When jukeboxes came, musicians said it would kill live music. Everything exists alongside each other now.”

The TikTok That Went Viral While I Slept

I told Patrick about my experience with viral content on another AI account. I made a humorous video about a wolf attacking a farmer in Drenthe. Posted it before taking a nap.

When I woke up, I had a WhatsApp from my mother-in-law with my own video. She didn’t even know I made it. It said “frequently forwarded” at the top.

That’s when I realized: timing and tapping into what’s currently happening is everything.

But here’s the challenge with BoxxBoxx. I made a video referencing a popular Formula 1 meme. It did nothing. Sometimes you hit, sometimes you miss. The only way to find the format people love is to try many things and learn from what works.

My Honest Doubt

Near the end of our drive, I laid out my real concern.

“Let’s say the song is actually good. It has hit potential. But people know she doesn’t exist. Can they still feel something? Can they become fans?”

Patrick paused. “That’s really the question.”

But then he circled back to Harry Potter, to Mickey Mouse, to animated characters in Frozen. “They can fall in love with a character. Especially because you’ve chosen a shared passion. She loves Formula 1 too. That puts her ahead already.”

He’s right that adults connect with fictional characters. But there’s a difference between a movie character and an influencer trying to build a following in real-time on social media.

That’s the experiment.

What I’m Taking Away

This conversation clarified something important: I’ve been so excited about the AI tools that I’ve been putting the cart before the horse.

The bottleneck is me not doing the foundational work.

I need to create a proper character profile for BoxxBoxx. Define exactly who she is, what she stands for, what makes her vulnerable, what makes her unique. Write it down. Make it consistent.

The music has to be genuinely good. Not just “good for AI music.” Actually good.

And I need collaborations. Getting in front of audiences that already care about Formula 1 is essential.

Patrick’s final advice stuck with me: “Make sure you’re having fun with it. People feel that. They sense when something is made just to score versus when someone made it because they enjoyed making it.”

That part, at least, I’ve got covered. This project combines five of my hobbies: music, Formula 1, AI, marketing, and podcasting. I’m having a blast.

Whether BoxxBoxx becomes the first AI influencer to truly break through in the F1 world or remains an interesting experiment, I’m learning more than I could have imagined.

And who knows? Maybe in a few years, Patrick will be right about those AI radio DJs. And maybe BoxxBoxx will be sitting at the table during race analysis broadcasts, projected as a hologram, responding in real-time.

The technology is already almost there.

Now I just need to make her someone worth listening to.