Why AI and Authenticity Are NOT Opposites with Molly Mahoney

When I started planning the Summer of Good AI, I knew exactly who I wanted to kick things off.

Molly Mahoney has been teaching AI since 2021, two full years before most people had heard of ChatGPT. She came from the stage, performer first, marketer second. That background gave her something a lot of AI educators don’t have. She understood voice before she ever wrote a prompt. What you say, how you pause, what you leave out. She carried that understanding into her work long before generative AI was a household term.

She built a program, built a community, and built a framework around what she calls the soul problem. That phrase alone might be the most honest diagnosis of why AI content fails that I’ve heard.

What performers know that most AI users don’t

Molly referenced Stanislavski, the acting theorist, and something he put at the very bottom of his chart on how to be a true actor. One large box. “Work on yourself.”

Her point: if you don’t do that inner work, if you don’t know who you are, what you sound like, what you actually believe, everything you produce becomes a performance. Audiences can tell. Readers can tell. AI can amplify your voice, but it cannot invent one for you.

That idea became the foundation of everything else she shared.

The soul problem, and why most people are skipping it

Here’s what Molly said that I’ve repeated more than once since we recorded: “AI needs data. It needs a brain. But the thing people miss is that our AI also needs soul. If we don’t weave that soul piece in, we get spam.”

Most people go straight to prompting. They open a tool, type something vague, and then wonder why the output sounds generic. Like it could have been written by anyone, for anyone, about anything.

That’s the soul problem.

The AI isn’t broken. It’s working exactly as expected. It’s filling in the blanks with what it’s seen most of. If you haven’t given it your voice, your vocabulary, your actual point of view, it defaults to a version of everyone else’s.

What infrastructure actually means

Molly described a meeting with a prospect who asked if AI could just build them a landing page. Her answer: yes, if you have the infrastructure in place.

She opened Claude in front of them and built a landing page in five minutes. Not because she’s exceptional with prompts. Because she walked in with three documents already built. A company details document. A brand document. A visual document. The AI wasn’t guessing. It knew what to say, how to say it, and what it should look like.

Most businesses skip that step and go straight to the tool.

I see it with clients too. The ones who say “I tried AI and it didn’t work” almost always mean they tried prompting without giving the tool anything real to work with. No voice document. No defined message. Sometimes five people in the same company answering “what do you do?” five different ways.

Add AI on top of that and you don’t get efficiency. You get five versions of unclear, moving faster.

Why weird wins

Molly is big on weird. She’s not talking about gimmicks.

The business case: you need something recognizable so the right people can find you. If your content could have been written by anyone, it will be found by no one. You need something specific enough that the right person lands on it and thinks, yes, that’s her.

But the deeper reason matters more. If you’re not showing up as who you actually are, you’re not living in your truth. And that catches up with you.

She called it a Trojan horse. She says she’s teaching people to build their business, but underneath it all, she’s helping them figure out who they are so they can show up and live a truer life.

That’s the right frame for AI too. Not “how do I use this tool,” but “what am I trying to say, and who am I saying it to?”

One person, not the neighborhood

Near the end of our conversation, Molly brought up Mr. Rogers. Which is a sentence I didn’t expect to write in a post about AI, but here we are.

When his show launched, he didn’t say “Hey, everybody in the neighborhood.” He said “Hey, neighbor.” Singular. One person.

That choice made the whole thing. People trusted him because he spoke to them like a real human on the other side of the camera, not an audience.

Molly calls this the Mr. Rogers conversation, and it maps directly to content creation. The more specifically you write for one person, the more broadly it lands. AI can produce a lot of content fast. It doesn’t automatically make that content feel personal. That part is still on you, and it starts with knowing exactly who you’re talking to.

Try this today

Before your next piece of AI-generated content, answer two questions:

What’s one phrase you use all the time that you’ve never formally written down? The thing clients repeat back to you. The way you explain your work that makes people feel like they finally get it.
Who is the one specific person this is for? Give them a name if it helps.

Put both answers in your prompt. Then compare the output to what you’d get without them.

That gap is the soul problem.

Listen to the Full Episode

You can hear the full conversation with Molly on A Good Pour wherever you listen to podcasts. She also has a free ideal client tool that goes well beyond demographics, covering the specific details that change how you write and who you attract. You can find it at molly.livegoodai.