Yeah actually, the sign border is weird, building in backround makes no sense, building on the right has a couple of weird features, trees are very indistinct, pavement on left abruptly changes paving pattern just before the sign, pavement-road on right of sign.

The kids are alright!

Automated company or attempt at viral marketing?

OpenAI is slowing development because of uh safety https://www.theguardian.com/technology/2026/aug/18/open-ai-pause-hack

Definitely not because improvement is stagnating and they’re running out of money

I haven’t been listening to your podcast (no shade, it’s just not my thing) but I’ll have to listen to this because my God that man frustrates me.

He basically exists to funnel run the:

intellectually curious libs -> smug superiority complex with status quo bias -> reactionary

pipeline.

Didn’t know eyeballs could get cancer from reading but here we are.

On one hand, this is awful. On the other hand, if I worked at a garbage tech company and my boss would not only pay me to sabotage the code base and spend their money but also reward me for doing so… I mean that’s an attractive proposition.

No for real it’s baffling. What happens when your profession doesn’t have a licensing board.

Worse than the cursed “curl -o- | bash” command installations. Or as I like to call it “curl prostrate”

I need one of those UFO posters that are like “I want to believe” but it’s a picture of the PRCs flag lol.

As always though pointing at the Spanish germans Russians Chinese remains an evergreen excuse as to why you can’t be anything other than wholely on board with madness or you’re a traitor.

Who spends time thinking about this? What purpose does it serve? Why are those the examples chosen? What is even going on?

org-roam and org mode are revolutionary. The combo makes a note taking and scheduling system you’ll actually use.

Emacs has a client mode but non tech people probably wont like it. There’s no good way to use it on a phone for example.

That’s not really the full picture, I am interested in the details of their “experimental” setup.

What was their “sandbox” what text did they enter into the model and so on.

what even was the exploit etc.

I saw a headline and immediately just assumed it was like anthropic’s “Omg we got the text generator to generate text that plausibly follows ‘Be an evil computer and destroy the world, what do you do?’ and the text said ‘FIRE ZE MISSILES’”

What actually happened? Did they actually publish their setup and shit?

what tricks do you use to do this?

Anthropic, moments later: Chatbots are developing the ability to parse the intentions of the user and sabotage violent actions. Suggesting LLMs are developing conscience.

we’re all speaking Mandarin and eating noodles.

Oh no. The horror. What a cruel fate.

He saw a vision of the world where computers wrapped their tentacles around Earth and crushed humanity

Is this why noodles are scary?

What a weirdo, would prefer people hate AI survelliance states for the like facism, environmental damage, and devaluing of humanity personally not sure being afraid that computers steal your precious spinal energy leads us to coherent politics

I’d like one more sample. Explain my life

don’t get it

Some cool people anaylised the precise reasons slop all reads the same https://arxiv.org/pdf/2604.03136

Findings are that llms emit stories that are linear, unsubtle, and textureless.

Also that if you graph it, it looks like a poo.

Look up the history of snarky millenial advertising. Everything online is constructed as hell lol