Facebook slop: the future of ad-supported media?

Niamh McIntyre, for The Guardian:

These [Facebook] accounts – and there are hundreds, possibly thousands of them – present themselves as the work of British patriots. In one typical, AI-generated video, a middle-aged man claims his local cafe “has stopped serving pork, bacon and sausages just to avoid offending people”. Another post from the same account includes a sepia-tinted set of images of Victorian London, mourning a time when the city “was English, first-world and beautiful”. Alongside this type of reactionary nostalgia, it’s not unusual to see memes that call Islam a “cancer”, decry Muslims praying in public as an “invasion of the west” or promote the “great replacement theory” (which claims that white populations are being deliberately replaced by non-white immigrants).

For the past seven months, I have been investigating who is really behind pages like these. The answer, it turns out, is often young, entrepreneurial men from south Asia. They tend to have zero interest in UK politics, but the content they create often boosts far-right talking points in Britain and contributes to the increasingly hostile atmosphere for immigrants and British Muslims.

Read the whole thing: Who’s behind the Facebook page posting hateful AI slop about the UK? The answer might lie in south Asia.

The most important part about these “social media content creators,” from the point of view of Meta and the other Big Tech companies, is their material is cheap. Generative AI, low wages, zero research or news-gathering expenses—there is no way that a legit news site can compete in raw seconds of attention per currency unit. And they don’t just work in politics. There’s probably even more of this kind of slop about health-related topics.

Big Tech wins when they can move more ad money to support cheap slop and misinformation instead of more expensive content.

The problem for advertisers, of course, is the halo effect. Advertising just works better in a more trustworthy context. Content is not a commodity. Check that link for sources—I just added another one since I put that post up. Social science research and marketing research are notoriously hard to replicate, but the halo effect, along with a reverse halo effect fortoxic content is one of those results that keeps replicating, no matter how inconvenient it is for the Big Tech companies that oligopolize the advertising market. Slop is just not a drop-in replacement for legit content as an advertising context.

Even though ad-supported content is not a commodity, the Big Tech companies of today are used to running a strategy that Joel Spolsky summed up asCommoditize your complements back in 2002, and that Bill Gurley wrote more about recently. For example, replace an expensive input like Sun servers with racks of generic PC servers, and replace expensive licensed software with open source. The biggest open source successes aren’t the acquisitions of Red Hat and other pure open source companies. The biggest impact has been internal. Google, and later Meta, answered the question “build or buy?” with neither, and chose topeer produce instead.

That peer production decision was, in many cases, a benign one. Big Tech “kernel teams” work on the Linux project, which regularly cranks out not a datacenter release and a mobile device OS release, but a general-purpose kernel release that even a low-budget developer can customize for almost anything you would need an OS for. The companies also support peer production of video codecs through theAlliance for Open Media. YouTube doesn’t have to pay for patent licenses—and as a side effect, neither does an independent filmmaker or the builder of a home media PC. Commoditizing the complement, when it works, causes a lot of the good stuff on the Internet. Imagine how dismal it would be to deploy even a basic database-backed CRUD application if open source databases such as ifPostgreSQL hadn’t been supported by earlier developers. And the process continues in more adjacent IT markets. Today,RISC-V is a solid replacement for ARM for many uses.

The place where commoditization breaks down, though, is where there’s no commodity. Facebook Slop Guy is not a drop-in replacement for a legit site, even if you ignore thepositive externalities of the latter. (For more on various theories on how advertising really works, or not, seeThe Anatomy of Humbug.)

The only way to fit Facebook slop (and crappy app store and search ads) into the same market niche as legit sites is, well,

Control the coinage and the courts—let the rabble have the rest. — The Padishah Emperor, in Dune by Frank Herbert

If the attribution cartel can control the standards for how advertising results are measured (and do the math in an obfuscated way, hidden from third-party checking) then all of a sudden Facebook Slop Guy isn’t just a commodity replacement for a legit site, he’s a better choice, with ROAS going up and to the right. That’s what I was on about inWhat Happens When The Attribution Cartel Meets Advertising’s Halo Effect?.

Commoditizing the complement is great when the commodity product is a viable replacement for the scarce, costly original. But the attribution cartel is trying to use the wrong tool for the job. There is some hope here, because commodification strategies don’t always work.

• Lindows, later Linspire, was a high-profile project to do a desktop operating system compatible with Microsoft Windows.

• Open Source Applications Foundation intended to produce an open-source, interoperable personal information management suite (like a replacement for Microsoft Outlook+Exchange)

Both of those went up against Microsoft, and there’s no Microsoft this time. Just a lot of companies of all sizes that are either in the ad-supported publishing business, or the advertising agency business. But the failures didn’t just fail just because of who the adversary was. A big part of their story arc was getting too much attention, too soon. Neither one was ready. The time for a commoditize the complement project to become a story is when it’s capable of successful niche projects, likeLinux print servers. If it gets too much ink too early, it can flame out. Fortunately, the attribution cartel is not ready. By W3C’s own standards, they have a bunch ofoutstanding issues.