Our Human Thoughts on AI

Our (Human) Thoughts on AI

And why most agencies are getting it wrong.

Every agency in our industry is racing to slap the word “AI” on their capabilities deck. For too many, that means asking teams to plug prompts into off-the-shelf models, calling the output “strategy” and billing it at full rate. The result is slop. The potential for real damage is tremendous. And clients are sure to notice the discrepancy.

AI is already supercharging our team’s productivity and impact, without compromising security or strategic thinking. Here is how we think about it.

A piano doesn’t make music. A musician does. A razor doesn’t give a great shave. A barber does. AI, wielded correctly, makes our team exponentially more effective. When treated as a substitute for thinking, it makes everything worse.

AI didn’t just change our toolkit. It forced us to rethink every single process in the agency — from how we identify the right reporters for a story to how we conduct deep research on a client’s world to how we monitor media coverage and online chatter in real time. We didn’t retrofit. We rebuilt.

We use AI to stress-test a message, identify gaps in an argument, sharpen a metaphor and push our thinking further. Think of it as the world’s greatest sparring partner — one that never runs out of patience and pushes us to get to a stronger answer. The judgment, though, is always ours.

We don’t put our team on open AI models and hope for the best. We operate a closed-loop system that ensures no client information is ever run through a vulnerable model.

Wherever discretion is required—and it typically is in our business—one of Miller Ink’s well-trained team members is making the call. Full stop. AI surfaces, synthesizes and accelerates. Our team decides, edits and owns the outcome. The buck stops with us, not a machine.

The greatest risk of AI isn’t that it hallucinates. It’s that it’s horrifically mediocre. We have built quality controls into every stage of our process to ensure that what we put in front of clients is the product of real thinking — not just the first plausible, run-of-the-mill thing a model generates. When AI-generated work goes out unchecked, it erodes the client’s trust in the thinking behind it. A boilerplate press release, a pitch indistinguishable from every other pitch, a memo padded with filler — these indicate that your PR team wasn’t really paying attention. In communications, credibility is the product. The perception that you’ve lost it accelerates fast. The moment a client reads something and stops believing it, you have lost the room. Recovering from that is often impossible

Our AI infrastructure isn’t a one-size-fits-all system. We build a dedicated AI environment for each client — one that understands their objectives, their audiences, their voice and their constraints. It operates within the context of their needs, not across the entire agency.

Across every process we run, we have engineered repeatable workflows that enforce quality, ensure consistency and serve as training tools for our team. Great workflows don’t drive slop. They codify what great work looks like.

We have invested significant time and resources into training our team on the right and wrong ways to use AI. Not just which tools to use — but how to think about using them, when to trust the output and when to push back. That judgment doesn’t come from a prompt. It comes from experience and expertise.

We’ve built AI agents that can be made available directly to our clients — so they can access and leverage the tools we’ve developed internally. Our advantage is meant to be their advantage.

A dedicated team within the agency is constantly evaluating new AI tools and figuring out how to apply them to our workflows. At the same time, every team member has a standing mandate: keep experimenting. Keep pushing. The moment we think we’ve figured it out is the moment we fall behind.

Contacts

Phone

Email

Our (Human) Thoughts on AI