The Unofficial Apple Weblog returns as plagiarized hyperslop

The Unofficial Apple Weblog was one of the great tech blogs of its age, crankin' out the poasts on all matters related to the tech giant and its lovable yet locked-down gear. Back in the Boing Boing Gadgets days, we must have linked to it every third day. It was eventually shut down in 2015 by whatever corporate equity hole ended up owning it, but it came back to life a few days ago, complete with the original contributors. Alas, the uncanny content was lifted from MacWorld and revised by an A.I. set to paraphrase, and the original contributors' names were being used without permission. It turns out that the domain was bought by a Hong Kong ad agency and immediately put to use as a content farm.

Christina Warren, who left a long career in tech journalism to join Microsoft and later GitHub as a developer advocate, shared screenshots of what was happening on Tuesday. In the images, you can see that Warren has apparently been writing new posts as of this July — even though she hasn't worked at TUAW since 2009, she confirms to The Verge.

Another screenshot showed Warren's name listed next to what appears to be an AI-generated photo and a generic bio, alongside a list of other former TUAW writers, including Brett Terpstra, Chris Rawson, and Chris Ullrich. All of the listed authors have had their photos replaced with AI-generated images, 404 Media reports, and many tell 404 that they have no involvement with the new website. AppleInsider confirmed that its author William Gallagher's name was inappropriately attached to content by TUAW's new owner as well.

Ad people buy dead websites because of the search authority they have. There's nothing new about this trick, but now AI can fill these sites with superficially convincing fresh content too, as in this case. A more subtle ploy is to backlink from the purchased domain to other sites, providing them with the dead site's undead SEO blessings. Google says it's cracking down on AI-driven SEO manipulations, but Wired recently posted an alarming story suggesting that it either favors AI content in search results or is simply losing the war against it.

I like the term "hyperslop" to refer to AI content that systematically confuses reality with things posed as reality (AI content that merely tries to achieve realism is also hyperreal, but interesting mostly for its failure aesthetics). It's not about the quality of the output, it's about plugging the right receptors; it works because of vulnerabilities, not capabilities. This uncanny zombie blog is a good example because it addresses a publishing medium the way that "Boomers Liking Impossible Kitchens" addresses a social one. It's a shot at rewriting the perceptions (and expenses) of reality that constrain the effectiveness of advertising. And it failed because no matter how well they nailed their technical implementation, they used the real names of real journalists who are still working the same beats, then plagiarized the contemporary content of their popular peers. They were so focused on trying to make the text indistinguishable from human-written copy that they forgot about the presence of actual human beings. They couldn't see what the dogshit tesseract looked like in the higher dimensions. Their slop is fine but their hyperslop fails.