Sound Thinking: Creative Effectiveness in an Age of AI
I’ve been watching the AI conversation in our industry with equal parts fascination and fatigue.
The fascination comes from genuine curiosity. I’ve spent most of my career at the intersection of science and creativity, and AI is arguably the most disruptive force to arrive there in my lifetime. The fatigue comes from the noise. The breathless proclamations. The existential hand-wringing. The premature eulogies for the creative class, followed almost immediately by prognostications heralding a new age of creative empowerment. The reality, as usual, is far more interesting, and more nuanced, than the headlines.
Putting AI’s creativity to the test
At Studio Resonate, rather than theorizing about AI’s role in creative work, we ran an experiment. We assembled three creative teams (human, AI, and hybrid), gave them a real brief, and put their output through three rounds of rigorous evaluation: a live pitch to a media agency, scoring by a panel of creative directors and audio professionals, and quantitative testing of their audio ads by a third-party research company. We also asked each team to keep a journal, capturing what they were learning along the way.
The results of that exercise shaped how we think about managing our human talent when AI is introduced into the creative workflow. When teams leaned on AI for iteration (generating variations, stress-testing ideas, accelerating production), it delivered real value. But when teams began outsourcing the deeper work of judgment (the original spark, the strategic insight, the emotional resonance that builds brands over time), the work fell short. Not because the technology isn’t impressive. It is. But impressiveness isn’t the point. Creativity, at its most effective, engages our emotions, shapes our perception, and encodes memories. Accomplishing those ends, it turns out, still requires human intentionality and accountability.
Our research into human versus AI voices sharpens this picture considerably. When we tested both in real advertising contexts, most listeners couldn’t reliably tell them apart: their detection rate was barely better than a coin flip. Yet human voices still outperformed AI voices by 24% on implicit brand attraction, a non-conscious measure of how listeners actually feel about the voice they’re hearing. More striking still: when listeners were merely told a voice was AI-generated (even when it was actually human), trust collapsed by 27% and positivity fell nearly 9%. The label alone did the damage. That anti-AI/pro-human bias has remained consistent across years of testing, despite rapid improvements in the technology.
Humans bring inherent, irreplaceable skill to the table
This finding has real implications for the growing pressure to disclose AI use in creative content. Transparency is a value worth protecting. When consumers are primed that an advertisement was created with an AI voice, they penalize the work, regardless of its quality. Efficiency may push us towards imagining a world where AI replaces a human voice, but effectiveness (and favorability) demands a different approach: use AI to augment human performance, not as a substitute for it.
These points aren’t meant to make a case against AI, but to make a case for human judgment. For the voice that carries lived experience into a room. For the writer who knows the difference between a checklist in a brief and an insight that reveals a human truth. For the creative director who can feel when something is almost right but won’t stop tweaking until it is. These aren’t soft skills. They’re the irreplaceable ones. The research keeps telling us the same thing: audiences respond to humanity. They trust it, they feel it, and even when they can’t consciously identify it, they reward it. The most strategic question a brand can ask right now isn’t how much AI they can use, but how much human intentionality they can preserve.
So where does that leave us?
AI is genuinely improving our ability to iterate. It’s accelerating timelines, opening up dynamic testing that would have been cost-prohibitive before, and surfacing possibilities that human teams might not have explored on their own. In an audio-first environment, we’re using it to rapidly prototype sonic concepts, stress-test ideas, synthesize volumes of research, and simplify administrative tasks, all areas where speed and scale matter.
Where AI consistently falls short is in work that requires intentionality. Strategic insight. Cultural attunement. The decision to take a risk that the data can’t fully justify. These remain irreducibly human responsibilities. When it comes to AI generated work, someone has to take responsibility for considering its impact on brand perception and consumer behavior.
AI can augment creativity, but humans should lead
Which brings me to what I’ve affectionately dubbed, “Maverick’s Maxim.” In Top Gun: Maverick, Pete “Maverick” Mitchell reminds a room full of elite pilots of a core truth: “It’s not the plane. It’s the pilot.” The tools matter. The prompts matter. The willingness to experiment matters. But none of it matters more than the “human in the loop” who decides whether the creative approach is strong enough to earn attention, forge connection, generate emotion, and encode memory, and who takes responsibility for whether or not the output achieves those goals.
AI doesn’t change what great creative work requires. The real threat to creativity is less about artificial intelligence and more about the gradual replacement of discernment with optimization, of intentionality with automation, and of human judgment with systems that can generate outputs but can’t own their meaning.
In an Age of AI, sound thinking is still your competitive advantage. And that advantage is decidedly human.
Steve Keller is Sonic Strategy Director at Studio Resonate, SiriusXM Media’s audio-first creative agency. He blends behavioral science and cultural intelligence with audio-first frameworks to help brands make sound decisions.