I Was a Hard No on AI. Here's What Changed My Mind.

By Amanda Shilling

Let me be the straight shooter I am. When AI really started making noise, my gut reaction was hell no.

That probably surprises people who know how I lead my team at MINT, or maybe it doesn’t. Because everything we do here is built on relationships. The work we do for clients is driven by original thinking, deep intentionality, and the kind of creative investment that actually means something. The idea of handing that over to a machine felt like a betrayal of everything I’d built.

So yeah. Hard no.

Image link
What Actually Shifted Things

One of our core values at MINT is Adaptability. And at some point, I had to get honest with myself about whether I was living that value, or hiding behind my skepticism.

I kept hearing the same thing everywhere: Your job isn’t going to be replaced by AI. It’s going to be replaced by someone who knows how to use AI well. And I didn’t want our team to be behind the times. More than that, I didn’t want my own stubbornness to be the thing that held us back.

So I started testing. Quietly, thoughtfully, the way I do most things.

How I Actually Use It (And How I Don't)

Here’s the thing people get wrong. AI isn’t doing the work for us. It’s making us faster at the work we were already doing well.

I use it as a brainstorming partner. I think out loud. Talking has always been how I get my best ideas out, and I can talk a lot faster than I can type. So if I can speak my thoughts, have them transcribed, and then have AI help organize them into something coherent and readable? That’s efficiency. That’s more time for the stuff that actually requires a human.

I also use it the way I used to use a thesaurus. Sometimes you’re this close to the right word, the right angle, the right framing, and you just need something to push you there. AI does that.

What I don’t do is let it replace the voice, the intention, or the relationships. If I see a post that was clearly just generated and pushed out without a human touch? That’s where I lose my mind a little. Platforms are already starting to flag that content. Audiences are already tuning it out. And clients who are paying for original, thoughtful creative work deserve better than that.

The Perception Problem Is Real

One thing I think about a lot is: if a client hears we’re using AI, their first thought might be “Well, then why am I paying this much?”

And that’s a fair question if AI is “doing everything.” But it’s not. We’re doing more with it, not less. The brands we build still take the same level of care and intentionality. The thinking still happens. AI just helps us execute faster and serve clients better in the process.

That framing matters. Transparency matters. I was open with my team about my own evolution on this, where my head was before, where it is now, and I think that honesty goes a long way. Lead by example. Share real use cases. Show people what ethical, high-integrity AI use actually looks like in practice.

Where I See This Going

I’ve started exploring AI agents,  both for our own workflows and potentially as tools we could build for clients. We want support to continue after a brand hand-off if that is the next step. We want support to extend the value of the work we’ve already done together, even when a client doesn’t have the budget to keep investing with us.

That’s the version of AI that excites me. Support, not replacement. Extension, not shortcut.

The Human Edge Isn't Going Anywhere

Here’s what I keep coming back to… The heart-led approach wins. Not in a fluffy way. In a give a damn way. People are already starting to pull back from AI-saturated content. They can feel when something is missing. And we want to be there when that happens, when they remember why the human element mattered.

AI will keep evolving. New platforms will come. I’ll keep testing, assessing, and deciding what actually earns a place in how we work. But the core of what makes MINT MINT? That’s not something any tool can replicate.

We’re human first. That’s not going to change.