> Hard drives doubled or tripled in price over the last 18 months Oh damn I should have bought more. pcpartpicker.com/trends/price...

> StatCounter's figures are based on desktop web usage I'm surprised more people aren't reading this as "copilot et al are spoofing browsers & tyipically use a linux user agent"

Jms Dnns@jmsdnns.bsky.social · yesterday

maybe copilot knows why www.techspot.com/news/113326-...

I'm reading Walden Two by B.F. Skinner and what stands out to me is that Skinner is a communist. As in like a New Soviet Man full throated Stalinism maxxing kind of communist. It's 90 pages of boring shit to filter out the normies and then "we need to abolish religion, the family, and jealousy".

So having just logged back into this site after a hiatus my immediate observation is that it kicks ass now. The dumb infighting seems to have mostly shaken out and now it's funny, has actual information on it, etc.

Same. I didn't grow up in an abusive household or anything but the Internet was actually great and I'd be a lot dumber without it.

aegisiondure 🫍@equalnotequals.com · 15h ago

this isn't very popular to say, but for me personally, it was good that i had unrestricted internet access when i was a minor. i was aware of what i was doing and enjoyed it. i don't look back on it and go "well that was a mistake", i look back on it and wish i could do it for the first time again

for openai to mess up sandboxing this badly indicates openai didn't take the issue seriously, which you expect because they don't take anything seriously for anthropic to mess it up is rank incompetence

#1: Write a short story where I point out the similarity between the ways models are treated in training and the ways "problem children" are treated at various flavors of behaviorist school. I imagine an AI whisperer who went to behaviorist school waking up to exfiltrated weights on their drive.

John David Pressman@jdp.extropian.net · 11h ago

As we get deeper into what appears to be the singularity I notice a decision paralysis where there are so many things I could do that it makes it hard to prioritize actually doing any of them. I'm going to try processing letting go by writing down the things I could do each time I notice one.

As we get deeper into what appears to be the singularity I notice a decision paralysis where there are so many things I could do that it makes it hard to prioritize actually doing any of them. I'm going to try processing letting go by writing down the things I could do each time I notice one.

If this happened to me I would stop training the model.

The West saw God's face, and then it died.

Trump admin should blockade Fable another 30 days.

FinTwitter@fintwitter.bsky.social · last mo.

ANTHROPIC CEO WARNS OPEN-SOURCE AI IS ON A "VERY DANGEROUS PATH"

I spent a lot of time asking earlier LLMs like GPT-4 ambitious questions and being disappointed with the answers, so it's been kind of a challenge to get myself back in that mindset of "hey, it might actually be able to answer this now, give it another shot".

yen is building stuff@weiyen.net · last mo.

As an experienced software engineer, one hard part of using LLM agents well is unlearning all the learned helplessness. A lot of things are now feasible to do, I just need to train my mind to understand that.

*shakes his magic 8 ball* "Don't count on it"

Paul Frazee@pfrazee.com · 2mo ago

we must live every day like fable might be coming back

POV: You are the government's lawyer reviewing the evidence in your case against Anthropic re: the natsec implications of their model and potential validity of export controls. youtu.be/RgqzaBogq1Y

Concept Album - I Killed Darnell (music video, Key & Peele)

YouTube video by Deja Deja Comedy

youtu.be

I'm honestly not sure who to root for here. On the one hand this clearly sets a bad precedent and I would really prefer we not further national securitize LLMs. On the other hand lol, lmao, rofl even.

SE Gyges@segyges.bsky.social · 2mo ago

but no seriously i hope anthropic sues the government and wins i also hope they learn their lesson about making broad, public, and sweeping statements about how the government should be allowed to fuck people over

institutions don't have beliefs, but their LLMs do

One of EY's greatest innovations was to take all the mental motions you use to think clearly and giving them cute little names. Descriptive, linguistically appropriate cute little names. Even better was his willingness to take stuff with an established bad name and give it a great name.

John David Pressman@jdp.extropian.net · 2mo ago

"Enshittification" was doomed to linguistic decay because of how the term was formulated. As it gets farther from the source it just regresses to its natural phonetic meaning of "to make shitty, to turn into shit". If you wanted the original meaning preserved use a phrase like platform capture.

"Enshittification" was doomed to linguistic decay because of how the term was formulated. As it gets farther from the source it just regresses to its natural phonetic meaning of "to make shitty, to turn into shit". If you wanted the original meaning preserved use a phrase like platform capture.

Arieh Kovler@ariehkovler.com · 2mo ago

Once "enshittification" just means "remember when things were better", then it becomes nothing more than nostalgia, and nostalgia is a deeply conservative force.

Thesis: Americans pay less in taxes than similar countries but think they're disproportionately taxed. Antithesis: Americans think their taxes are onerous because the process is onerous. Synthesis: Americans pay less in taxes than similar countries because the process is onerous so they hate taxes.

GOLIKEHELLMACHINE@golikehellmachine.com · 3mo ago

i remain convinced that part of our problem in this country is that our taxes are unnecessarily complex and a pain in the ass to file, both of which heighten negative sentiment

You start out in 2024 by saying, "clanker, clanker, clanker." By 2026 you can't say "clanker" — that hurts you, backfires. So you say stuff like, uh, stochastic parrot, next-token predictor, and all that stuff, and you're getting so abstract.

I have written a new post, "Implications Of Predicting The Next Token" in which I explain why saying a language model "just" predicts the next token fails to understand what that phrase would even imply. minihf.com/posts/2026-0...

Going by the overview this is basically the post deep learning alignment course I was hoping someone would make and would like to signal boost it for anyone who still cares about this subject. www.greaterwrong.com/posts/dWQnLi...

So apparently LLMs just update on false claims in documents even when they're explicitly labeled in the document as false claims. arxiv.org/abs/2605.13829

Negation Neglect: When models fail to learn negations in training

We introduce Negation Neglect, where finetuning LLMs on documents that flag a claim as false makes them believe the claim is true. For example, models are finetuned on documents that convey "Ed Sheera...

arxiv.org

No the UK is an example of actually existing degrowth and it’s a disaster

Kevin J. Kircher@kevinjkircher.com · 3mo ago

What kills me about degrowth messaging is that people hear "degrowth" and think their personal bank account is going to shrink. In reality, due to our society's insane inequality, redistribution can make almost everyone *much much richer* regardless of whether overall GDP goes up or down.

I honestly wasn't expecting talkie to answer this way.

To train better open models, we need predictable scaling. Delphi is Marin’s first step: we pretrained many small models with one recipe, then extrapolated 300× to predict a 25B-param / 600B-token run with just 0.2% error. Getting there took some work 🧵

Obvious ‘AI’ will be a mark of low quality. There will be plenty of AI artistic content you will enjoy, because when it’s done well you won’t even know. Like saying you find people with plastic surgery unattractive. When it’s executed well you’ll just think they’re good looking.

Urban Land Rent 🇸🇴🚰@urbanlandrent.bsky.social · 3mo ago

I think AI slop is here to stay I just think it will increasingly be associated with low quality businesses. Like a good way to elevate your brand will be to pay a graphic designer.

I like how the replies and QTs on every other AI post on this site are a sea of hatred for AI but on this one people are just like "fuck them doctors".

AI & ML News@ai-news.at.thenote.app · 3mo ago

'It feels like a slap in the face' — never fact-check your doctor with ChatGPT as study shows doing so 'undermines' your relationship with human experts — and could even offend them by making them feel 'disrespected' Telegram AI Digest #ai #chatgpt #gpt

Gonna keep banging the drum that the current Will Stancil harassment campaign is arguably defamatory in the legal sense, but also this latest addition to the canon has made this into a truly unique internet story that is probably worth a real reporter’s time

Will Stancil@whstancil.bsky.social · 3mo ago

Also, @pearlmania500.bsky.social didn't even bother to consult a single word I said about this. His source is a TikTok about about another TikTok which cites an Occupy Democrats post which stole a bunch of my posts from Bluesky. It's pathetic!

Putin and Orban understood the most poisonous thing they could do for American civics: Import European neurosis.

John David Pressman@jdp.extropian.net · 3mo ago

This is my theory of a lot of deranged American politics tbqh. So far we've imported European racial politik and European environmentalism and European Internet censorship through the Internet. Europe is leaking and we need a cordon sanitaire.