AI Didn’t Make Me Faster. It Made Me Better.

It was 4:12 a.m. and I was still working. Not because a deadline demanded it or a client was waiting. Because AI is deepening my dedication to the craft of making ads. 

Most of the discussion about AI in advertising and communications is centered around efficiency and speed. I use AI in the ruthless pursuit of excellence. My new creative process is labor intensive – achingly so.

I’ve spent the last 10 months immersing myself in all the major LLMs. Most agencies in politics are using AI to do what they already do, just faster. Some seasoned operatives argue the immediate value of AI is its ability to offload administrative burdens, so content creators can be freed to think deeper thoughts. I’ll argue that both of those mindsets miss the true potential of AI. For me, AI is about deploying intensive creative discipline into every aspect of making an ad, from ideation to scripting to editing and testing. 

Prior to infusing AI into my creative process, I could crank out a client-approvable script in about five minutes. AI has added significant hours to my work week. I am working longer and harder than ever before because of AI, and I’m loving every minute of it. I’m more invigorated professionally today than when I was working on my first statewide political campaigns in Michigan three decades ago. 

As a creative director, I have a creative temperament. I’ve never particularly enjoyed collaborative brainstorming sessions. I’ve too often bristled at feedback on my ads. Those are my weaknesses. With AI, I don’t have those issues. I can take the criticism or feedback from the machine more constructively than from a colleague. That new dynamic has pushed me to elevate my craft in a profound way. 

With my ego fully in check, I am much more willing to edit, refine and push boundaries. I have no fear of thinking out loud and receiving a blank disapproving stare from colleagues. If the AI flags a flaw in my thinking, we fix it together without judgment in the pursuit of a higher standard. That’s the biggest difference AI’s made for me. It’s also making me far more open to receiving feedback from others. In small and big ways, AI is helping me grow deep into my career when old habits are hard to change.

Getting the most out of AI, means rolling up your sleeves and grinding. A willingness to iterate, re-grade and iterate again until achieving excellence is hard work. There are no shortcuts in the AI process I deploy. I want our industry to understand that the best AI is led by human inspiration and active involvement. While some may attempt to duplicate my exhaustive AI process, my results will always be unique to me because they are driven by my creativity and experience.

Here’s the multi-step AI methodology that’s keeping me grinding through the night. On big ads and campaigns, I run every script through Google’s Gemini. I have trained Gemini on 125 of my award-winning ads. It knows my voice. It knows the advertising the marketplace rewards with big statewide wins and prestigious awards. It’s a powerful ideation and creative tool that is infused with the DNA of the very best of my ad-making process. I iterate against Gemini until landing a perfect A grade on a script or message progression. 

I am open-eyed about the tendency of AI to flatter its users. That’s why I run the same script through Anthropic’s Claude to get another set of AI eyes on my work. I instruct both models to grade me honestly. I’m not looking for my ego to be fluffed. I want FP1 to win campaigns and awards. Ensuring that two separate models trained differently give a concept an impeccable grade is a tough standard to meet. Only the best creative survives this process. On really important projects, I take the process a couple steps deeper by running red-team simulated focus groups of AI personas. 

One of my favorite innovations is creating a two-persona AI panel of the archetypal male and female voters my candidate needs to win and running ad concepts and messaging exercises against them. I then have Nano Banana – Google’s AI image-generation tool – create an image of these male and female voters, so I can visualize who I’m speaking to when I write ads directed to them. The archetypes come with detailed biographical notes on marital status, occupations and hometowns. With AI, I can now clearly picture the exact voters I need to persuade to win the election. The insights I’m gaining from these synthetic voter panels rival the best focus groups I’ve ever sat in. I’ll never write a message progression again for a big campaign without kicking the tires first with a red team synthetic panel.

In the span of a baseball season, AI has helped me rewire the entire way I make ads. I’m seeing the results every day of my new creative roadmap. The technology has expanded my ambitions as an ad maker.

Earlier this year, as the Texas Republican Senate primary heated up, FP1 cut an ad for Sen. John Cornyn’s re-election campaign. Rep. Wesley Hunt (R-Texas) was picking up steam in the primary. We needed to blunt his momentum. I hoped to do something memorable that played off Hunt’s name and mocked his physique. I landed on the creative hook of “That Dog Won’t Hunt.”  My initial idea involved Hunt with a fat and sleepy basset hound. After a couple of iterations with Gemini, we landed on pairing Hunt with a Pomeranian – a tiny showdog that undermined Hunt’s tough guy image.

Beyond ideation and scripting, AI opens creative doors that were previously shut because of cost prohibitive shoots. Rather than casting a Hunt look-alike actor and conducting a $100,000 shoot that would have taken weeks to plan and execute, I was able to create an AI likeness of Hunt and produce the ad in six hours. Before we ran “Show Dog,” we pressure tested it against synthetic focus group panels of Republican primary voters in the Dallas and Houston suburbs.

Yes, it was a labor-intensive process. But the results speak for themselves.

“Show Dog” struck an important balance that I feel is imperative when deploying AI in political ads. While it used a likeness of Hunt, it avoided the dreaded deepfake label, because it was clearly a parody ad. It was genuinely funny. People watch the ad and they focus on the creative hook of the Pomeranian and the “Won’t Hunt” idiom. They see the craft and hear the message. They don’t fixate on the AI. That was the true achievement.

My experience grinding inside the LLMs suggests that the best uses of AI will be executed by hardworking practitioners who drive the AI. While I understand the value and the efficiency of agents, I find my hands-on process that requires my active participation with the AI at every juncture to be superior. AI shouldn’t be a speed tool, but a quality enhancer.

Here is the paradox with AI. The most innovative users of the technology will find it adds time to their workflows. When you see greater possibilities in front of you, you become more ambitious. Beyond using AI as a writing partner and editing tool, I am spending hours experimenting with synthetic audiences, because I think they are going to become the new frontier of AI in politics. The ability to red team an ad with AI personas in real time has been revelatory for me as a content creator. Synthetic testing is the future, and I want to be at the cutting edge of this innovation. 

Trent Wisecup is a Partner and the Creative Director at FP1 Strategies and PLUS Communications.