AI Can’t Quilt: Why Human Craft Still Matters in the Age of AI

An introspective look at creativity, artificial intelligence, and how to use powerful new tools without surrendering what makes the work human.

I recently had lunch with one of the founders of Adventure House, someone I hadn’t seen in several years. The agency was founded in 1990, and I bought it in 2012. Over that time, it has survived enormous changes in technology, media, client expectations, and the way creative work gets produced.

Yet, for the most part, the process hasn’t really changed: thoughtful collaboration between people, and a willingness to keep innovating and pushing creativity forward to get results.

At its best, agency work means listening carefully, asking the hard questions, interpreting what is and what isn’t being said, challenging assumptions, and sometimes struggling with an idea until something better emerges (the messy, uncomfortable part). The tools that support that process keep evolving, yet the process itself remains deeply human, at least until we find ourselves marketing to robots.

The lunch left me thinking about the conundrum of leading a creative agency at a time when artificial intelligence is reshaping nearly every part of our industry.

The founder is now spending a portion of his retirement focusing on creative endeavors, including honing the art of quilting, with his work recently featured in a national quilting publication. There’s something so profoundly human about the handcrafted pursuit of quilting: choosing the materials, developing a composition, working through mistakes, building something slowly, and getting better through repetition.

AI can't quilt. At least, it can't physically experience what it means to quilt.

A machine can generate a pattern, but it can’t understand the patience, frustration, curiosity, and satisfaction of producing something original through years of practice, sweat, and learning. And that distinction matters far beyond quilting.

I’ve watched the same tension play out in music. When I’m not running the agency, I’m usually holding a guitar. I’ve been playing since I was 10 or 12, and lately the guitar community has been consumed by its own debate about AI and creative authenticity: AI-generated songs, AI-cloned YouTube channels, and automated production tools raising hard questions about what listeners are hearing and what companies are presenting as authentically human.

AI can have a legitimate place in music production, whether for speeding up tasks, experimenting with an arrangement, or handling technical work that once took much more time. But when a company is selling a product made for musicians, it should portray that product’s performance honestly.

A recent headline involving D’Addario, a guitar string manufacturer, is a useful case study. A highly polished video showcasing its new guitar strings sounded artificial enough that listeners began questioning whether AI had played a role in creating the music. The company said that the track was written and performed by a person, but later acknowledged that AI-assisted tools had been used during the mixing and mastering process.*

By then, a delayed explanation, deleted comments, and a response video that raised even more questions had complicated the story. The audience no longer knew what to trust, and that is a lesson all agencies and brands should take seriously. Once that circle of trust is broken, it is difficult to get back.

People can sense when something is “off”

People can’t always prove that a piece of writing, an image, or a whole campaign came from AI. But they can often sense when something feels off.

The language is polished but generic. The image is impressive but odd or lifeless. The campaign looks like several others. The brand voice doesn’t reflect the company. Everything is technically correct, and yet it feels just… off.

AI didn’t invent generic work. Agencies were producing bland, forgettable campaigns long before ChatGPT or Claude. What AI has done is make generic faster and easier and put it into the hands of everyone. It can generate hundreds of meh headlines, summarize hours of research, write code, and produce variations faster than any human.

But infinite output isn’t the same as a great idea.

The temptation, then, is to use AI mainly to save time, effort, and cost. Our responsibility is to use the time it frees up to push harder and produce better work.

The real skill is knowing which question to ask. Which answer to challenge. Which idea to develop, and which to discard. And most important: when do we stop generating and start doing the hard part?

AI can offer up (plenty of) options, but it can’t take responsibility for the best choice. That remains our job.

 

What AI means inside the agency

AI is a crutch, a sidekick, a wall to bounce ideas off, the fastest and smartest, yet sometimes seems to fall asleep at the wheel.

AI’s value depends on how thoughtfully it’s used. We use it for research, to organize complex information, test ideas, troubleshoot technical issues, speed up repetitive tasks, and produce initial drafts that help establish direction. But that’s only the beginning, not the end.

The danger is when initial drafts become the finished work. A first draft isn’t a strategy. An attractive image isn’t a visual identity. A collection of facts isn’t a unique insight. A website assembled quickly isn’t necessarily one built around a company’s customers, goals, content, technology, and long-term needs. AI helps us move faster; it doesn’t eliminate the hard parts or the thinking the work requires.

Like everyone else, we’re still fine-tuning our workflow as the technology evolves, but a few principles guide us through the AI creative conundrum:

  1. Creativity and a fresh POV still lead. Everything we create should be crafted for the brand and give its audience a reason to believe, trust, or learn something new. AI can generate possibilities; people decide which one actually resonates.

  2. Brands don’t break through by prompts alone. It takes clear thinking, strong ideas, and branded messaging that is uniquely their own. It’s in all the details a project or creative brief doesn’t cover: endless questions, trial and error, not hitting it on the first take, picking up on a key message that opens up a whole new direction, and bringing it all together in a fresh, creative package. 

  3. Authenticity is sacred. People want something real behind a company: its history, expertise, people, stories, and process. AI can smooth those details away; it takes human direction to put the specificity back in.

  4. Honesty matters. Using a tool to research, build processes, organize notes, or test a headline is different from presenting a fabricated performance or an AI-generated testimonial as real. 

  5. Human review isn’t optional. From art direction and writing to technical implementation, facts are checked, language is crafted, technical best practices are followed, and brand standards are applied. Someone is always accountable at every step.

Always push harder.

Perhaps the real conundrum is that while AI provides limitless possibilities and helps streamline processes, the best work still depends on the same things it always has: listening, questioning, refining, and knowing when something isn’t good enough.

That work begins with client conversations, discovery, research, and the creative brief. It continues as we decide what has potential, what adds something new, what needs to be pushed further, and what belongs in the ‘circular file’.

The tools will keep evolving, but the responsibility to create something original, honest, and meaningful remains in our hands.

Whether it’s a video, digital campaign, a website, and, yes, even a quilt, that is still the craft.

* D’Addario sets the record straight on AI