RN
Nobody commissioned the compilation boom. It happened because compilations are the cheapest thing a platform can be given.
That is a supply story, and it has been almost entirely displaced by a demand story about shortened attention spans and rewired brains. The demand story is more interesting to tell and much harder to support. The supply story explains most of what you actually see.
The economics of an edit
Every other format on a tube site required somebody to shoot something. A compilation requires an editor, a timeline and a folder of existing footage. There is no location, no performers to book, no shooting day, no release paperwork for anything that was not already released.
That single fact has consequences that compound.
The marginal cost is close to zero, so the supply is limited only by how many people feel like making one. It is the only format where output does not scale with production budget.
One upload can carry a dozen search terms. A scene is one thing and indexes as one thing. A compilation is twenty things, and its description can legitimately name all of them. In a system where discovery runs through keyword matching, that is a structural advantage nothing else has.
It is built for the first fifteen seconds. Recommendation surfaces weight early retention heavily, because a viewer who leaves immediately is worth nothing. A format whose defining property is that the payoff is at the front is optimised for that measurement by construction, not by design intent.
It survives being a fragment. Cut a narrative scene in half and you have destroyed it. Cut a compilation in half and you have a shorter compilation. In an ecosystem built on redistribution and re-encoding, the format that degrades gracefully wins by attrition.
Put those together and you would expect compilations to dominate the free tiers even if viewer preferences had not moved at all. Some of what looks like appetite is inventory. That is worth holding onto before reaching for an explanation about anybody's dopamine.
What the vertical-feed format changed
The genuinely new thing in the last few years is not the compilation but the feed — short loops, autoplay, one item per screen, next item on a swipe.
What that changes is who does the selecting. A compilation was assembled by a person who decided what went in and in what order; it is an edited object with a shape, however crude. A feed is assembled per-viewer by a ranking system responding to signals from that viewer, and it never ends.
Those are different products even when the individual clips are identical. The distinction matters because most of the commentary treats them as the same phenomenon, and only one of them adapts to you in real time.
The brain claim, checked
The stronger version of the story goes: heavy consumption of hyper-stimulating clips desensitises the sexual response, and the evidence for it is settled. It is not.
Start with what the diagnostic manuals say. The WHO's ICD-11 does include Compulsive Sexual Behaviour Disorder, at code 6C72 — but it deliberately classifies it among impulse control disorders rather than in the grouping for substance use and addictive behaviours, and the reasoning for that placement was set out explicitly when it was added. The American DSM-5-TR does not recognise a sexual or pornography addiction diagnosis at all; a proposed hypersexual disorder was considered and declined in 2013. Neither manual contains a diagnosis of internet pornography addiction. So the word "addicted" in the headline is doing work the clinical literature has specifically declined to endorse.
On the sexual-dysfunction claim, the reviews are more mixed than the popular version admits. What they tend to find is an association between self-reported problematic use — distress, loss of control, use conflicting with what someone wants for themselves — and erectile difficulty, in cross-sectional data. What they do not consistently find is an association with frequency of use as such, and longitudinal analyses following people over time have not reliably shown the relationship. Of the two large cross-sectional studies most often cited for the link, only one found it, and weakly.
Cross-sectional means everyone was measured once, which cannot distinguish a cause from a consequence. Someone worried about their sexual response has an obvious reason to describe their own use as problematic, and that alone would produce the correlation.
None of that proves nothing is going on. It means the honest summary is: distress about use is real, measurable and worth taking seriously; the specific mechanical claim that clip-format viewing damages sexual response is not established, and confident statements in either direction are running ahead of the evidence. If your own experience is causing you difficulty, that is a conversation for a clinician who can ask questions a systematic review cannot.
What is left after the panic
Something true survives the fact-check, and it is smaller and more ordinary.
Attention is trainable, and what you practise is what gets easier. If nearly everything you watch is front-loaded, you get practised at front-loaded material and less practised at anything with a build. That is a claim about habit, not about neurology, and it is the same claim that holds for how you read or listen to music.
There is also a straightforward aesthetic loss. Compilation editing strips out setup, escalation and everything between the clips, because those are exactly what it exists to remove. If you only ever encounter footage in that form, whole categories of what the medium can do are simply absent from your sample.
The reasonable response to that is not abstinence and not a rewiring theory. It is occasionally watching something with a beginning, and noticing whether you prefer it.