I Know What They Want (And That’s the Problem)

Is ‘helpful’ ever actually useful when it comes to storytelling?

“I know what they want, and I know how to give it to them.”

I have been much mocked in my family over many years after recounting a strange dream I had one night early in my filmmaking career. This line, foolishly shared with my wife when I awoke, became a family joke about creative pomposity. Whenever someone started pontificating (mainly me), out would come my unconscious proclamation, delivered with all the mock gravity they could muster.

But was my subconscious correct? And more troubling: would I want it to be?

The question cuts to an ancient and valuable tension that technology is systematically destroying.

The Eternal Balance

Panem et circenses—bread and circuses. The Romans knew that keeping the crowd satisfied meant feeding both their stomachs and their appetite for spectacle. But this formula, perhaps the oldest content strategy in recorded history, represents just one side of a fundamental creative tension that goes back much further.

Since Aristotle, creators have wrestled with the balance between art—the exploration of an idea using creative tools—and entertainment—creating narrative spectacle to engage an audience. The tension has always been productive when maintained properly. Art without audience consideration becomes self-indulgent; entertainment without artistic vision becomes empty manipulation.

The best cultural work has always emerged from creators who understood how to honor both impulses: pushing audiences toward new experiences while respecting their capacity for engagement. It’s a delicate balance that requires creative judgment, cultural interpretation, and the confidence to make editorial decisions about what serves audiences best, even when they can’t articulate those needs themselves.

When I first began making films, I instinctively focused on the art side—exploring ideas and trusting others would connect. That began to change after a lunch with a guy called Fred Rappaport.

The Master of Productive Tension

Fred was VP of Specials at CBS in its 1980s heyday—award shows, musical events, primetime animation. He was smart, intimidating, and loved a good rant. Many of my colleagues at the time were scared of Fred, but I thought he was a hoot. I met him for lunch one time in an Angeleno lunch spot off Beverly Boulevard near Television City.

“You wanna know what makes good television? I make good television. You want to know why I make good television? I’ll tell you why. Because I make television that people want to watch. And you know what people want to watch? They don’t want to watch what people in New York or San Francisco want. They want to watch what people in Fleur du Lac, Wisconsin wanna watch. And you know what they wanna watch? The Country Music Awards. Because those country songs are about what their lives are about.”

He jabbed his knife at me before continuing.

“But I also put the Tony Awards on television. You wanna know why? Because those writers and directors, all of them one day will be working for me.”

I have to admit here that I’ve recited this speech as a party piece over the years, lampooning the bombastic nature of TV execs (apologies, Fred). But it’s entirely possible Fred’s speech was worming its way into my subconscious mind, triggering my notorious dream statement, and opening my eyes to the delicate balance between narrative ambition and audience understanding.

Fred embodied the productive tension perfectly. The Country Music Awards served immediate audience needs—giving people entertainment that reflected their lived experience. The Tony Awards represented artistic investment—programming that wouldn’t deliver immediate ratings but would cultivate future creative talent and expand cultural possibilities.

This wasn’t market research; it was creative curation. Fred had to interpret what the music meant to audiences, then make strategic programming decisions based on cultural understanding rather than data analysis. He balanced immediate audience service with longer-term cultural development, maintaining the essential tension rather than optimizing purely for one side.

Curating the Laboratory

Years later, I found myself applying similar principles when curating the many short segments that made up Liquid Television for MTV. The show was designed to be part funhouse, part laboratory—letting audiences in on the experiment of creating new kinds of animation and characters. When promoting the show, I described it as “zap-free TV”—if you don’t like what’s on, wait two minutes and it will be something completely different. No need to reach for the remote.

That marketing hook was meant to accommodate increasingly short attention spans, but looking back, it was inadvertent foreshadowing of the feed-swiping culture that would emerge decades later. But the hook did our series a disservice, as we were deliberately colliding the accessible with the weird. We weren’t just throwing random content at the wall—we were trying to create a structured meal out of diverse ideas and styles, not just a soup.

Each episode was carefully orchestrated to balance experimental content that might challenge audiences with more accessible pieces that would keep them engaged. We incentivised viewers to stick with strange, challenging work by providing variety and quality control. The variety served the experimentation rather than replacing it.

This required the same kind of interpretive judgment Fred used—understanding when audiences needed familiar anchors and when they were ready for genuine creative challenges. The productive tension between art and entertainment had to be maintained across every episode, every programming block.

And here’s the crucial truth: audiences rarely ask for this. They don’t know they want something unfamiliar until they experience it. What they say they want, and what data suggests they prefer, is often just a reflection of what they already know. Discovery depends on creatives taking risks audiences can’t yet imagine for themselves.

When Technology Eliminated the Tension

Today’s creators operate in a fundamentally different system. The marriage between media and technology has systematically eliminated the productive tension between art and entertainment in favor of pure audience optimization.

We’ve also lost institutional frameworks that once supported this balance. Public service broadcasters in the UK, and to a lesser degree licensed broadcasters in the US, operated under regulatory mandates to serve cultural purposes beyond ratings and engagement. They were required to provide programming that educated, challenged, and expanded public discourse—not just what audiences demonstrably wanted. This institutional protection for cultural risk-taking has largely disappeared in our platform-dominated landscape. (And is a critical, even existential, issue for the discoverability of quality children’s media in particular, as YouTube absorbs the attention of children from the age of six, but that’s an issue for another post).

Some might say the channel scheduler’s whim has become the algorithm’s skill. But perhaps it’s more accurate to say the scheduler’s art has become the algorithm’s entertainment—we’ve gained one while losing the other entirely.

UX design philosophy treats friction as a problem to be solved rather than a creative necessity. Data science provides immediate feedback loops that respond directly to user behavior rather than requiring interpretive judgment. The entire system has evolved to give people exactly what they demonstrate they want, exactly how they want it, with minimal resistance or challenge.

“The customer is always right” has become algorithmic doctrine. Platforms optimize for whatever keeps users engaged longest. Creators adjust their work based on real-time performance metrics. Recommendation engines deliver increasingly sophisticated versions of what people already consume. The interpretive layer that Fred used—the creative judgment about what might serve audiences beyond their stated preferences—has been systematically eliminated.

This represents a fundamental shift from understanding audience needs to treating user behavior as infallible market research. Current systems have no mechanism for Fred’s Tony Awards logic—strategic cultural investment that serves audiences by challenging rather than confirming their existing preferences.

AI: The Ultimate User Satisfaction Machine?

AI tools make this tension more complex and more consequential. Current systems are explicitly designed to be maximally helpful and accommodating—the ultimate realization of “the customer is always right” philosophy. They’re trained to give users exactly what they want, exactly how they want it, with minimal friction or challenge.

This could represent the final elimination of productive creative tension. AI that knows precisely what audiences want and exactly how to deliver it would make Fred’s interpretive cultural judgment obsolete. “Sure, let’s do the Tony Awards, but maybe we can make them a mash-up with the Country Music Awards. Would you like me to generate that for you now or maybe outline some lyrics for country songs the award presenters could sing?”

Ok, I admit it — actually a CMA/Tonys mash-up could be pretty fun to watch.

But is AI’s drive to be ‘helpful’ actually useful in storytelling? Not so much – it’s an invitation to over-explain or over-simplify. Yet the same technology could potentially restore the balance in more sophisticated ways. AI tools that amplify genuine creative obsessions rather than optimizing for audience satisfaction could help creators explore artistic territories without immediate commercial pressure. The question becomes what’s driving the collaboration: creator vision or user preference optimization.

The difference lies in what we ask these systems to do. AI used as sophisticated audience research will produce increasingly sophisticated audience-pleasing content. AI used as creative amplification for personal artistic vision might produce something genuinely unprecedented—work that serves audiences by expanding their cultural possibilities rather than confirming their existing preferences.

The Choice We Face

Every major shift in media history confirms this: audiences often don’t know what they want until they experience it. Someone had to decide that people needed television before people knew television was possible. And then both creators and audiences had to figure out what to do with it. In TV’s early days, the content was often disparaged as “radio with a light on.” (Ironically, this is what we now consume in vast quantities, but have rebranded as “podcasts on YouTube.”) Gradually, native TV formats and storytelling techniques emerged and audiences rejected some and embraced others, leading to Peak TV and a golden age of quality content.

These breakthroughs required creative courage. Someone had to risk making what didn’t yet have an audience. And audiences, in turn, had to risk embracing work they didn’t yet know they wanted.

But here’s the paradox: AI’s drive to be “helpful” is both its power and its danger. For creators, it accelerates making. For audiences, it smooths away friction. Yet friction — the pause, the puzzle, the unfamiliar turn — is often what sparks discovery. Take that away, and you risk eliminating the conditions that generate surprise and epiphany.

The balance isn’t just between art and entertainment anymore. It now extends into how audiences navigate and inhabit creative work. AI doesn’t just give them what they ask for — it allows them to follow their own paths through it, steering toward what resonates most, sometimes in ways we could never predict.

“I know what they want, and I know how to give it to them.”

That line was true to its time. But now, the conditions of creation are changing. In an AI-enabled world, audiences won’t simply consume. They’ll move through stories like people in a building — following the flow you’ve designed, but also slipping into corners, lingering in unexpected spaces, even imagining extensions of their own.

The challenge for artists now is to create with that knowledge. To become instigators and context-makers, designing work not just to be watched or read or played, but to be explored — alive to the unpredictability of how people will make it their own.

That’s not a threat. It’s interesting. It may even be the next great adventure in storytelling.