The cliché runs: pornography decides every format war. VHS beat Betamax because the adult industry chose VHS. The internet was built by porn. Every new medium is pioneered by smut and then gentrified by everyone else. It is repeated with the particular confidence people reserve for facts they have never checked, usually by men, usually at the third drink, usually in the tone of someone letting you in on the real story behind the official story.
Like most such claims it is a mixture of the true, the unfalsifiable and the flatly wrong, and it is worth separating the strands, because the wrong parts have been doing damage to clear thinking for forty years.
Start with the videotape, since it is the load-bearing example. The story is that Sony, in a fit of corporate prudishness, refused to license Betamax for adult content, the adult industry went to VHS, and consumers followed the pornography. It is a marvellous story. It has the shape of a parable — the puritan punished by the market — and it has the additional charm of implying that the real decisions are made by the disreputable. The trouble is that the format’s fate can be explained perfectly well without it. Betamax’s original cassettes ran for an hour; VHS ran for two, and two hours holds a feature film and, more importantly for the mass market, an American football game. JVC licensed aggressively and cheaply to a wide circle of manufacturers; Sony licensed narrowly and dearly. The machines were cheaper. The tapes were cheaper. The rental shops, which are what actually determined what people watched at home, stocked what they could get most of for the least money. Adult tapes were on both formats. They were on Betamax in quantity. The claim that Sony banned them is a claim in search of a document.
What is true — and interesting, and usually skipped over in the rush to the joke — is a set of much less cinematic facts. The French videotex system Minitel, distributed free to millions of households from the early 1980s, generated a substantial and famous proportion of its traffic and revenue from the messageries roses, the pink message services, which were essentially paid flirtation by terminal. Premium-rate telephony in Britain and America was built commercially on the back of chat lines. The early commercial web’s payment plumbing — recurring billing, chargeback management, the whole miserable apparatus of proving to a bank that a transaction is legitimate — was worked out first and most thoroughly by people selling access to pictures. Affiliate marketing, in the form we now recognise on every corner of the internet, was refined in the adult trade before it was respectable. Streaming video at scale, with adaptive bitrates and cheap delivery, was solved commercially by tube sites years before the arrival of prestige television made it a subject of polite conversation.
So the honest version is not ‘porn wins format wars’. It is something more specific and more useful. Sexual demand is unusually early, unusually price-insensitive at the margin, and unusually willing to tolerate a bad experience in exchange for access. People will squint at a postage-stamp video, wait four minutes for a JPEG to unfurl line by line down a fourteen-inch monitor, install unsigned software, hand a credit card to a company in a jurisdiction they cannot spell. This is why sexual users arrive first. It is not evidence that they decide who wins.
from Infinite Desire (2026)
Claude Fable 5.1:
The passage’s central move is right, and I think it can be sharpened and corrected in a couple of places.
The videotape section. The debunking is sound, but one detail undercuts it: two hours does not hold an American football game, which runs about three. The actual history is better for the author’s case than what’s written. RCA declined to go with Betamax partly over recording time, then pressed Matsushita for a four-hour mode specifically so a football game would fit, and sold its 1977 VHS machines on that figure. So the football point belongs to the four-hour LP mode, not the two-hour cassette. The section could also make the structural point that a format licensor has no practical veto over content anyway. Anyone with a duplication bank could put anything on Beta, and did. The strongest surviving version of the legend is that Sony was unhelpful about duplication through its own facilities, which is thinly sourced and, even if true, isn’t a ban.
The best falsifier, though, isn’t in the passage at all. When the cliché was dusted off for Blu-ray versus HD DVD, much of the adult industry leaned toward HD DVD, which was cheaper to replicate, and the same Sony-is-prudish rumour circulated. Blu-ray won anyway, decided by the PlayStation 3 and Warner’s defection. That is the one time the claim made a testable prediction in advance, and it failed.
The ‘true’ list. Most of it holds, with gradations. Minitel is right, though the messageries’ share of traffic is routinely inflated. It peaked in the mid-1980s and declined, and the phone directory was always the most-used service. The more durable innovation was the Kiosque billing model, in which the carrier bills the user and splits revenue with the service provider, essentially the app store thirty years early. Payments is the strongest item. Affiliate marketing is fair as phrased (‘refined’, not ‘invented’), since flower sellers and Amazon have rival origin claims. Streaming is the weakest. Tube sites arrived after YouTube, ran on plain Flash progressive download for years, and adaptive bitrate came out of Move Networks, Microsoft and Apple. The better adult-industry example is a decade earlier: live paid video streams in the mid-1990s, when nobody else could make anyone pay for video online at all.
The thesis. ‘Early, price-insensitive, tolerant of bad experience’ is good, but it omits the property that explains the most: privacy. Sexual demand doesn’t chase novelty as such. It chases whatever reduces the social cost of access. The VCR moved consumption from a cinema to a living room, and the internet moved it from a shop counter to a bedroom. This accounts for the pattern the passage is after. Porn matters between media, where the privacy gain is large, and is irrelevant between formats within a medium, where Beta and VHS are equally private. Adult titles were a huge fraction of early prerecorded tape sales, which helped sell VCRs generally and tells you nothing about which VCR.
A second omission: much of the adult trade’s innovation was forced by exclusion rather than chosen. It mastered chargebacks because banks classed it as high-risk and because its customers had an unusual tendency to deny their own purchases. That reframes ‘pioneered then gentrified’ as something less flattering to both sides. Platforms tolerate sexual content while they need volume and expel it when they need advertisers or payment processors, as Tumblr, the OnlyFans wobble and the card networks’ treatment of Pornhub all show. The pioneers don’t get gentrified alongside the neighbourhood. They get evicted from it.
Two notes on the writing. The ‘third drink’ sociology is enjoyable but does no argumentative work, and it slightly licenses the reader to dismiss the claim for who says it rather than for why it’s wrong. And the excerpt promises three strands (true, unfalsifiable, flatly wrong) but delivers only two. ‘The internet was built by porn’ is presumably the unfalsifiable one, and it deserves its own treatment: nobody can specify what share of early bandwidth, revenue or engineering effort would make it true or false, and that vagueness is why the claim survives.
Given the 2026 date, the obvious test of the refined thesis is generative AI, where sexual and companion uses arrived early, tolerated poor quality and paid readily, and have had no evident say over which labs or architectures prevail. If the book goes there, the thesis predicts the outcome well.
