Much Profit, But Little Prophet

Years ago, I was talking with a creative director at The Richards Group, the first ad agency I worked for. He’d just come back from judging a creative awards show and shared an observation that I’ve never forgotten.

One of the agencies in the running was Crispin Porter + Bogusky, who had been doing headline grabbing work for Burger King. Until that point, the ambition of advertisers was essentially artistic. You wrote the best scripts. You hired the best directors. You invested in production. You made something beautiful that might last for months, maybe even years.

Crispin threw that standard out the window. They hired young, cheap directors. They worked fast and lean. They made spots in volume. Burger King’s work might run for a few weeks and then get replaced by something new. It wasn’t about making the perfect thing. It was about making imperfect things in volume.

Looking back, I think Crispin Porter + Bogusky may have been among the first mainstream purveyors of what we now call slop.

Content Becomes King

The phrase “content is king” is older than most people realize. Bill Gates wrote an essay with that title in 1996, predicting that the internet would become a marketplace for content. But it took another decade and a half for that prophecy to fully land. By the early 2010s, brands had internalized the lesson: publish constantly, feed the algorithm, chase the click (A/B testing anyone?). Volume became strategy. The goal wasn’t to make something great. It was to make something. And then make something else.

Content was still hard to make, though. Writing took time. Video took crews. Even mediocre production had friction built in. That friction served a quiet purpose. It was the last natural speed bump between an idea and the world.

Then AI arrived and removed that friction entirely.

Flood the Zone

Ted Gioia, writing in The Honest Broker, has a name for what happened next. He calls it “Flood the Zone,” an aesthetic process that now dominates the creative economy. Turn art into content for monetization. Dump everything on the market. Accept that the result is slop.

Gioia traces it across music, film, publishing, and visual art.

• AI startup Suno generates a hundred million songs every two weeks.

• Amazon is flooded with AI-generated books pretending to be written by humans.

• DeviantArt galleries, once a home for working artists, have been overrun.

• A single scammer in North Carolina recently pleaded guilty to stealing ten million dollars from human musicians by flooding streaming platforms with AI tracks and promoting them with bots.

It’s not just happening in entertainment. Hollywood has its own structural version of slop, built into the economics of intellectual property. Studios that purchase rights to major franchises operate under use-it-or-lose-it agreements. They must produce franchise content on a schedule or risk losing the rights. So they produce. Not because they have a story worth telling, but because the clock is running. The result is a Marvel universe stretched past coherence, reboots nobody asked for, sequels made by obligation.

And Now, Market Research

I’ve been watching this disrupt my own field of market research. And I think we just hit peak insanity.

Anthropic, the company behind the AI model Claude, recently published the results of what they’re calling the largest qualitative study ever conducted. Over one week, their AI interviewer spoke with 80,508 people across 159 countries and 70 languages, asking them about their hopes and concerns around artificial intelligence.

The methodology is genuinely impressive. The scale is unprecedented. But I keep coming back to a question the study doesn’t quite answer: why 81,000?

No one in their right mind would have ever considered conducting 81,000 qualitative interviews before AI made it possible. The reason wasn’t budget or ambition. It was judgment. Qualitative researchers have always understood that depth trumps volume in their discipline. The best insights come from the pauses. From letting silence do some work. From the moment a respondent contradicts themselves and you lean in.

An AI interviewer doesn’t lean in. It moves to the next question. And then to the next respondent.

As Ted Gioia might put it, there is much profit in selling the ability to conduct 81,000 interviews. But little prophet.

What Are We Giving Up?

Scale is seductive. It feels like progress. More interviews, more data, more confidence in the findings. But qualitative research was never designed to be scaled. The whole point was to sit with someone long enough that they told you something true, something they didn’t know they thought until the conversation gave them room to find it.

When we optimize for volume, we don’t just change the method. We change what we’re capable of discovering. The unexpected turn. The respondent who reframes the entire research question with a single offhand comment. The insight nobody thought to code for. These things don’t survive at scale. They require a human being paying close attention.

The market research industry has always prided itself on rigor. The question worth posing now is simple: if the goal is genuine understanding, is more better? And if the answer is no, what are we actually selling when we lean into slop?