Source: in-cyprus.philenews.com
One of the first articles I wrote for this column, at the start of the year, was about the “Dead Internet Theory.” Today I thought I’d check in on how that theory is doing.
The occasion was an article I read about the film Gods Don’t Give Gifts, which I suppose will be translated into Greek as Οι Θεοί Δεν Κάνουν Δώρα. It is the first film made entirely by AI that will be shown in US cinemas in December. According to Variety, the studio chose to delay the premiere, hoping to land Oscar nominations. In that case, it is obvious that many other creators will follow the same path, with everything that entails for the film industry as a whole. The trailer is on YouTube, and I recommend you look it up and judge for yourselves. Personally, I found it disgusting, and I wholeheartedly hope it fails.
I find the impressive advance of generative AI on the internet more worrying. Allow me to introduce Tiffany Sloan, an Instagram account that opened on September 11 and already has over 400,000 followers. A glance at the photo in the article is enough to tell you which audience has embraced her so warmly. The persona is a stand-up comedian who happens to have an OnlyFans account, for anyone interested in “getting to know her better.” On October 3, comedian George Brett Olson set one of her jokes beside a performance of his own: the same text, the same setting, almost the same outfit. Olson has in fact found other accounts that have stolen his jokes in the same way, with personas of artificial women to suit every taste.
Then there is Jean Philanthrope. He appeared on Instagram in September, in a brown suit, with a blond bob and a curled mustache, and began shadowboxing in stranger and stranger places. He quickly passed 228,000 followers, and the cryptocurrency that bears his name, $JEANPHIL, reached a market cap of $12.2 million before it collapsed. A few days later Archibald Brown appeared, also made by a machine, with a cryptocurrency of his own. People who do not exist acquire an audience, and the audience becomes a market before anyone has time to ask who is hiding behind it.
Some people fall easily for accounts like these, but the great majority of the public seems to be pushing back against the flood of AI content, which now looks inevitable. A Wall Street Journal poll from late September finds that 63% of adults in the US want an immediate freeze on the development of artificial intelligence. A Politico survey says that about two-thirds have at least “moderate” anxiety that the technology might destroy humanity. OpenAI and Anthropic are preparing for every eventuality. They are planning their response to a possible mass cyberattack, or to a political revolt after a catastrophe caused by their models. OpenAI calls these “preparedness exercises” and clarifies that the scenarios “are not treated as inevitable.” When you are preparing for the day after a revolt against you, though, the wording is not reassuring. Anthropic, according to Futurism, is already setting up a surveillance apparatus for the protests, on a “pre-crime” logic.
The conclusion is that the internet as we knew it has indeed died, and is being replaced, ever more quickly, by artificial intelligence. I hope I am wrong, but an internet that applauds Tiffany Sloan and Jean Philanthrope does not leave me much room for optimism.