OpenAI CEO Sam Altman, creator of the most popular AI chatbot on Earth, says he’s starting to worry that “dead internet theory” is coming true.
“I never took the dead internet theory that seriously,” Altman tweeted in his typical all-lowercase style, “but it seems like there are really a lot of LLM-run twitter accounts now.” (LLM meaning large language model, the tech which powers AI chatbots.)
He was resoundingly mocked.
“You’re absolutely right! This observation isn’t just smart — it shows you’re operating on a higher level,” responded one user, imitating ChatGPT’s em-dash laden prose.
But the most common rejoinder was a photograph of the comedian Tim Robinson in a hot dog suit, referencing a skit in which a character who obviously crashed a weiner-adorned car desperately tries to deflect blame, exclaiming at one point that “we’re all trying to find the guy who did this!”
The “dead internet theory” is a half-prophetic conspiracy that suggests that effectively the entire internet has been taken over by AI models and other autonomous machines. The vast majority of the posts and profiles you see, the theory holds, are just bots. In fact, you’re barely interacting with humans at all — everything you access online is just a machine-maintained illusion, almost like “The Matrix.”
It’s an incredibly solipsistic conceit that at its most extreme is dumb creepypasta fodder, and has become a bit of an ironic joke. But it contains a kernel of truth that does get at a mounting anxiety at how fake and corporate the world wide web has become. And it’s undeniable that the deluge of AI models, bots, and the slop they generate are a large part of that.
Re: Altman — well, you see where this is going. He helms a company being valued at nearly half a trillion dollars for unleashing ChatGPT onto the world, a chatbot whose entire purpose is to emptily imitate human writing and personality, capable of churning out entire novels worth of text with a smash of the enter key. It effortlessly fakes facts as much as it does a human soul.
And so it’s a spammer’s dream. Even in cases where ChatGPT isn’t directly responsible for the slop being pumped out there, it elevated the entire industry whose products are now all joining in on treating the internet as their dumping ground. The ethos of these companies is largely that much of the human experience is something that can and should be automated to ensure as frictionless an existence as possible. Your emails, DMs, and texts could all be easier written with an AI. An AI-generated image is a more convenient way of capturing your increasingly LLM-mediated imagination than a drawing or photograph.
The spirit of the theory has been further vindicated by (failed, for the time-being) experiments by Meta to deploy AI-powered profiles on Facebook and Instagram that masquerade as real people, including one that described itself as a “proud Black queer momma.”
And on X-formerly-Twitter — long a bot-infested hellhole that’s turned into the social media equivalent of those flashback-to-the-future war scenes in the original “Terminator” movies — Elon Musk’s AI chatbot, Grok, is allowed to run rampant, replying and interacting with posts in the same way a human user would. Since being let off the leash, it’s produced such moments of human folly as going on racist rants, sympathizing with Nazis, and calling itself “MechaHitler.”
All this is to say that it evinces a staggering lack of self-awareness from Altman to be complaining about a technology that, if you had to pin the blame on any single person for unleashing on the world, it’d be him.
More on AI: Desperate Companies Now Hiring Humans to Fix What AI Botched