April 10, 2026
Paging Dr. Make‑Believe
Scientists invented a fake disease. AI told people it was real
Chatbots fell for a fake illness, and the crowd is yelling “garbage in, garbage out”
TLDR: A researcher invented a fake eye disease with jokey clues, yet chatbots and even some papers treated it as real. Commenters are split between “garbage in, garbage out,” fears of people gaming AI advice (like product picks), and blame on sloppy publishing—raising big questions about who and what we trust online.
Move over WebMD panic — meet bixonimania, a totally invented eye “disease” dreamed up by researcher Almira Osmanovic Thunström, complete with fake preprints and links, a made-up author, and Easter eggs like thank-yous to Starfleet and the Sideshow Bob Foundation. The twist? Major chatbots started repeating it as real, and even some peer-reviewed papers cited it. The internet’s reaction: part outrage, part facepalm, part meme-fest.
The loudest chorus is pure “garbage in, garbage out” energy. One user shrugged that of course bots will parrot whatever’s online if there’s no counter-story, warning that bad actors can seed the web and bend AI advice their way. That ignited a bigger fear: if a fake illness can spread, what stops influencers from gaming AI shopping tips? Commenters confessed they’ve already bought products based on AI suggestions and are now sweating how easy it might be to rig rankings.
But another camp says the real villain is the publishing pipeline, not just the bots. If preprint servers and journals don’t catch obvious jokes about the USS Enterprise, why should machines? Meanwhile, jokesters nominated the stunt for an Ig Nobel, dunked on the name “bixonimania,” and posted “Paging Dr. Make-Believe” memes. Drama score: high. Trust score: low. And the community is split between laughing, panicking, and checking their eyeballs in the mirror.
Key Points
- •A Swedish research team led by Almira Osmanovic Thunström invented a fake condition, “bixonimania,” to test AI susceptibility to misinformation.
- •They seeded the term with Medium posts (15 March 2024) and uploaded two spoof preprints to SciProfiles (26 April and 6 May 2024).
- •The fabricated papers contained obvious clues of fakery, including fictional authors, institutions, and explicit statements that the study was made up.
- •Within weeks, popular AI chatbots, including Bing’s Copilot, began describing the fake condition as real.
- •The fake preprints were later cited in peer-reviewed literature, suggesting some researchers may rely on AI-generated citations without verifying sources.