John 🇺🇦

Mad Scientist. Really mad. At you.

[John 🇺🇦 @udarnik]

OK, Twitter. I stayed here for Ukraine content, to subvert Elmo. However, I can't in good conscience continue to support Elmo's business with even my tiny bit of content.

This will be my last post, here.

I'm the same handle on BlueSky:

@ https://t.co/mv09Z44OY7

[John 🇺🇦 @udarnik]

@EPICGOPFAIL @ashlarblocks Could be. Logistics and PMCS are foreign concepts to most of the Russian military. But they still have some juice left in the tank, so I'm not counting chickens just yet.

[John 🇺🇦 @udarnik]

@ashlarblocks @EPICGOPFAIL Gains from the beginning of the war, specifically around Kerson. I suspect a tactically stupid decision might be made to retreat to Krym. Russian generals aren't going to contradict Volodya about that.

[John 🇺🇦 @udarnik]

@EPICGOPFAIL @ashlarblocks A lot depends on whether those earth berms they're building can replace the pontoon bridges, and how they stand up to fire. I think the timelines are up in the air for now, depending on how much Russia decides to throw at the problem.

[John 🇺🇦 @udarnik]

@EPICGOPFAIL @ashlarblocks If they wait, they will *definitely* be walking. Are there many pack animals left in Krym?

[John 🇺🇦 @udarnik]

@EPICGOPFAIL One of the less dim ones.

[John 🇺🇦 @udarnik]

@AmyOwl73 @petite_michelle That container is just begging for some Diet Coke and a couple of Mentos.

[John 🇺🇦 @udarnik]

@ashlarblocks @EPICGOPFAIL Yes. Russia will start abandoning other gains to reinforce Krim, because they don't want to lose face by losing the peninsula. That will slow things a bit.

[John 🇺🇦 @udarnik]

@EPICGOPFAIL Many will bug out early, but I bet we're going to see a number of *really* slow Ruzzians making a run for the border just before it becomes too late to do so.

[John 🇺🇦 @udarnik]

@sameBZee @EPICGOPFAIL I absolutely love Ніч Упаде.

https://t.co/UUFovRyHuO

[Regina Bauer 🇪🇪🇺🇦 @petite_michelle]

A russian asked a very dumb question on Threads:
When the SMO ends, how many years do you think it will take us to be able to travel to Ukraine as regular tourists?

An Ukrainian replied: To come - yes, you will be able to do so, but come back from here - no way.

You'll meet a woman on the platform who buried her husband. A taxi driver who was wounded, a saleswoman whose son has gone missing... In short, the policeman who'll be looking for you also has combat experience... He won't find you.

[John 🇺🇦 @udarnik]

@An0nTheDoomer @SpaghettiKozak Boils down to sullen resentment at their treatment by their own leaders, wrapped in an unsubstantiated superiority complex they've been spoon fed about their language, culture and religion, seasoned with cognitive dissonance when confronted with the truth.

Some mystery.

[John 🇺🇦 @udarnik]

@sameBZee @EPICGOPFAIL Did you see Цвинтар dropped a new mini-album?

https://t.co/KSwa1zD81Q

[John 🇺🇦 @udarnik]

@sameBZee @EPICGOPFAIL Leningrad begs to differ:

"В Москве охуительно нюхать"

https://t.co/MiMuyHyFww

[John 🇺🇦 @udarnik]

@EPICGOPFAIL @speckledmoth As always, there is a Leningrad song for this:

https://t.co/NgqF0epJ7m

[John 🇺🇦 @udarnik]

@EPICGOPFAIL

Did you see this?

https://t.co/qEIf3SjMTE

[Nav Toor @heynavtoor]

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.

[John 🇺🇦 @udarnik]

@EPICGOPFAIL He meant nanometers.

[John 🇺🇦 @udarnik]

@EPICGOPFAIL And, as always, there is a Leningrad song for this occasion:

https://t.co/85DChPVMRY

[Eve Fairbanks @evefairbanks]

This is the most perfectly uncanny piece of AI-assisted writing I've seen--and a great tutorial on how to identify AI text. The OP is a captain! But the tell for AI isn't rhythm, wording, or fact errors. It’s that problems with *all these elements* exist equally & at once. (1/7)