Primaries. Vote in the goddamn motherfucking primaries. Get these lick-spittle, piece of shit, controlled-opposition weasel dicks out.

Every time I have to use my work laptop (with Windows) for anything, it feels like a giant step back. Lately it’s even worse; it feels like that step is right into some dog shit.

This might legitimately be the year of the Linux desktop, not because Linux suddenly got better, but because Windows finally got unacceptably bad.

We should be clear: They weren’t really anti-war. What they were was stupid enough to believe Trump.

A very cynical part of me wants to know who their parents voted for.

One possibility is that with Israel focused on Iran and Gaza, other countries take it as an opportunity to attack Israel. The U.S. escalates in defense of Israel, other countries escalate and start including U.S. bases as targets (becoming defacto allies of Iran in the process), and it all spirals from there.

I was going to say all this, only worse and less detailed. Thanks to you, I don’t have to!

My favorite example: Copilot to read your emails for you and send responses automatically. Get two people with Copilot sending each other emails with neither person actually involved. Efficient!

Tall guy, here. There are specific exercises you can do to strengthen and stabilize your back, and that might help. I’ve actually got fewer back troubles than I did when I was young and stupid, because I’m taking care of it.

How about a $10 billion fine for OpenAI for every mistake? Make it hurt. Make them pull the plug on this travesty.

Weird way to spell “pragmatic and effective.”

This is what failing to enforce their orders has wrought for the judicial branch. ICE will continue ignoring them until judges start ordering their bailiffs to take these scumbags straight to jail from the courtroom.

Thanks!

It goes beyond the problems introduced by the model router, though. I have to work with GPT 5.2 for my job (along with Claude, Gemini, and a few others), and we have enterprise API access to it. So when I select GPT 5.2 as the model to use, it’s spending tokens to actually use it.

And it’s pretty bad. It’s noticeably worse than the 4.x series. I find myself having to fix its mistakes far more often.

I’ve struggled to reason out an explanation, and model collapse really seems like a contender, especially if you follow information theory and why training these things is so hard.

As it happens, there’s a new talk about exactly this from George D. Montañez. You might find it interesting: https://youtu.be/ShusuVq32hc

Then why are newer versions of the major models performing so poorly? For instance, GPT 5.2 is definitely not an improvement over 4.5. What’s the root cause?

And that is why I no longer buy anything from them. I’m just embarrassed it took me as long as it did to realize what they were really doing.

I mean, we’re watching it happen. I don’t think it’s hypothetical anymore.

It’s already happening. GPT 5.2 is noticeably worse than previous versions.

It’s called model collapse.

Oh, the guy from Hermit Tech! He’s great, and his blog is hilarious and poignant in turns (though sometimes both at the same time).

You wanna know who really bags on LLMs? Actual AI developers. I work with some, and you’ve never heard someone shit all over this garbage like someone who works with neural networks for a living.