You have noticed it. ChatGPT feels dumber than it used to. Your prompts that worked six months ago produce worse results now. The writing sounds flatter. The ideas sound safer. The internet itself feels like it is shrinking. Every article reads the same. Every email sounds the same. Every answer sounds like it was written by the same voice.
You thought it was you. It is not you.
Researchers at Oxford and Cambridge published a paper in Nature proving what is happening. They call it Model Collapse.
Here is the mechanism in one sentence. AI trained on AI-generated data gets dumber every generation until it forgets what real human data looked like.
The internet is filling with AI-generated content. Blog posts. Articles. Reviews. Comments. Social media. AI companies scrape the internet to train the next generation of models. Which means the next generation of AI is being trained on the output of the current generation.
Each cycle loses information. Not randomly. It loses the rarest, most unusual, most creative parts first. The researchers call these the "tails of the distribution." The weird ideas. The unexpected perspectives. The things that made the internet feel human. Those disappear first.
What remains is the average. The safe. The expected. The bland.
Then the next generation trains on that. And loses more. And the next generation trains on that. And loses more. The researchers proved this is not a slow decline. Major degradation happens within just a few iterations. Even when some of the original human data is preserved.
They tested it on large language models. On image generators. On statistical models. The pattern was the same every time. The output converges toward a narrow, flattened version of reality that looks nothing like the original data.
The lead researcher put it plainly. "Large language models are like fire. A useful tool. But one that pollutes the environment."
The pollution is invisible. You cannot see which sentence on the internet was written by a human and which was written by AI. Neither can the AI that is about to train on it. And once the tails are gone, they do not come back. The damage is irreversible.
This is not a prediction anymore. It is a diagnosis.
The internet you grew up on was built by humans writing things no algorithm would have written. Strange, personal, imperfect, alive. That internet is being diluted. One generation of AI at a time. And the models trained on what remains are learning a smaller and smaller version of the world.
Model Collapse is not a technical problem. It is a cultural one. The thing that made the internet worth reading is the thing that disappears first.
that scans and makes sense.Clio the cat, ?July 1997-1 May 2016 Kira the cat, ??2010-3 August 2018 Jasper the Ruffian cat ???-4 November 2021 Georgina the cat ?2006-4 December 2025 Toni the cat ?2005-25 March 2026
...that a tweet lamenting the curse of AI should be written in the 'It's not x; it's y' ChatGPTese I've complained about here before:
'You thought it was you. It is not you. [...] This is not a prediction anymore. It is a diagnosis [...] Model Collapse is not a technical problem. It is a cultural one'
Unless the tweet was written by AI, in which case the joke's on us.
Anyway, there are probably negative ramifications that I'm not considering, but over all this seems like good news, with the AI bubble eating itself in only a couple of years. If all the tech bros, investors and security spooks have to eat shit and get a real job, so much the better. Though as ever I expect they'll find a way to turn their problem into everybody's problem. Good thing there weren't problems like climate change, species extinction, peak oil etc screaming out for investment and policy change at the time they were f*ing around with this, eh?
Posted by Ken Waldron on June 11, 2026, 7:49 pm, in reply to "Ironic..."
"...the 'It's not x; it's y' "
Takes me back to when my boys were young and I told one of them: "...Pick up your dolly from the floor..." The indignant rebuttal came quicky: "...It's NOT a DOLLY. It's a "Boys Action Figure"..."
To the extent they’re actually empowering, then they would presumably be deliberately enshitified ..
If AI were useful as a generally accessible and empowering took, it would inevitably be deliberately algorithmically enshitified after the current, shining honeymoon adoption trap.
Nothing in the world is provided to empower people in general. It will empower the few, presumably by first addicting us.
on June 11, 2026, 8:37 am
