Something looked off last month when I was scrolling the recommendation row on one of the big tube sites. The algorithm knows me well enough at this point. But the row felt different in a way I could not immediately explain. One slot was a generative video, the kind where the lighting is slightly wrong and the skin texture is almost right but not quite. Two slots over was a shaky phone video shot in what looked like a college apartment, no intro card, no music sting, just two people and a desk lamp.And sitting in the middle of that same row, completely untouched by the click data, was a glossy 4K studio scene with a professionally lit set, two performers clearly following a shot list, and a title that had been A/B tested by somebody in marketing. The studio scene had fewer clicks than either of the other two. I spent the next week pulling category data and what I found was not subtle.
The Two Things That Are Actually Growing

What Is Losing, and I Am Going to Be Specific About It
The scripted studio middle is down around 18 percent month-over-month in engagement across the categories I track, and the decline has not stabilized. When I say the scripted middle, I mean an aesthetic instantly recognizable to anyone who has spent time on a tube site in the last decade. The massage scene. The fake taxi, fake hospital, fake Uber, fake handyman, fake whatever-premise-the-writer-landed-on-that-week. The step-family scenario filmed in a production house in the San Fernando Valley with furniture arranged to suggest a generic suburban home. The glossy 4K studio set with three-point lighting and two performers hitting their marks with the energy of someone filling out paperwork.This aesthetic dominated tube sites from roughly 2010 to 2022. It was not a niche. It was the default. The thing every studio was producing, every aggregator was hosting, and every recommendation engine was surfacing because it was the thing that existed in volume. The massage scene was not a joke category. It was a load-bearing pillar of the industry for over a decade. And now it is losing audience from both directions simultaneously, which is a specific kind of pressure the industry does not have great vocabulary for yet.
Why Both Ends Are Winning for the Exact Same Reason

The One Thing No Production Budget Can Buy
Amateur content delivers the same feeling through a completely different mechanism. The messy apartment. The bad camera angle. The moment where the audio drops because someone bumped the phone. The sense that nobody rehearsed this, nobody is watching the clock, and nobody is thinking about how this plays on a 4K television. All of that signals “this was not engineered for an audience,” which paradoxically makes it feel more personal than anything a studio produces. The viewer is not the target demographic. The viewer is an accidental witness. That feeling of accidentally witnessing something real is worth a lot, and it is something no production budget can authentically manufacture.The scripted middle cannot do either of those things. It is too polished to feel real and too generic to feel custom. A massage scene with a named performer and a named studio and a professionally lit set sends one message above everything else: this content was produced for a mass audience, and you are the mass audience. That is the one message viewers have decided, with their clicks, they no longer want to receive.
What This Does to the Studios
The economics here are not complicated and they are not kind. Studios that built their entire production model around the scripted-middle aesthetic are getting squeezed from two directions at once, and neither direction is going to relent. AI undercuts them on cost by a margin that is not close. Generative video does not carry performer day rates, location costs, crew overhead, post-production labor, or distribution infrastructure at the scale a physical studio carries. The cost per minute of AI-generated content is dropping every quarter and that compression is structural, not cyclical.Amateur undercuts them on authenticity, which was the other thing the scripted middle used to sell. The implicit case for watching a professionally produced massage scene over a phone-shot homemade video used to be that you were watching skilled performers executing a well-realized scenario. That case holds for the niche that specifically values craft and performance. For the general audience it no longer does, because the general audience has concluded that the more produced something looks, the more alienating the experience becomes. The craft itself signals the scene was manufactured for a generic viewer, and that signal is now lethal to engagement.The franchise series that dominated 2015 to 2020, the themed-series model, the big tube network in-house studio output that filled recommendation rows for years: all of it is now pinned between a technology that undercuts it on personalization and a category that undercuts it on credibility. There is no lane left in the middle that the middle actually owns. Studios built around this model are not going to pivot fast enough to AI production because their existing infrastructure requires the pipeline to keep running to cover its own costs. And they cannot pivot to amateur because their brand is the opposite of amateur. They are structurally stuck in the thing that is losing.

What the Audience Is Actually Communicating

What Happens Over the Next 18 Months
The scripted-studio aesthetic will not disappear. I want to be direct about that because the easy takes are both wrong. “Studios will adapt” is filler with no specific claim in it. “Studios will die” overcalls it. What will happen is that the scripted middle shrinks into a specialty niche, the same way old-school 1970s loop-format porn became a niche after VHS changed the production economics in the 1980s. The form does not go away. It gets smaller, it gets claimed by the audience that specifically wants it, and it stops being the default content type filling every recommendation row on every major platform.There is a version of the scripted studio aesthetic that survives as a deliberate quality tier. The same way some viewers specifically seek out shot-on-film independent cinema in a streaming era because the format itself carries meaning for them, there will be an audience for high-production-value scripted content that values craft and performance. That audience exists and it will keep the form alive. But it stops being mainstream. It becomes the thing you actively seek out rather than the thing that gets surfaced to you by default.On the AI side, the number I am watching is not content volume, which will keep climbing regardless of what else happens. It is the share of AI-generated content coming from individual user prompts versus studio-scale batch generation. Right now the majority is still being produced at scale by distributors running it through existing channels. As the tools get cheaper and more accessible, that share shifts toward individual generation. When it does, the “made for you” feeling stops being a sales pitch and starts being literally accurate for a much larger portion of the audience.

The Read That Ties It Together



