I do often feel like if the documentation were better, this wouldn't be necessary, but sometimes the culprit turns out to be interactions between things or competition for some finite resource that even very good documentation would not foresee.
I am pretty skeptical, but I have been surprised by AI (specifically Google Gemini)'s powers of debugging. In my computing world, it's relatively common to try a thing and get a weird error message and an unmanageable trace-back. LLMs are a *lot* better than I am at parsing these.
This is not a shitpost, I'm serious. Adding things is so easy that any computer can do it. Determining whether it improves the whole (and leaving the whole intact when removing the excess) is purely human, because it requires a point of view.
This is just a shitpost, but there's something to the idea that
Almost every design choice and optimization for large language model-based "AI" comes from the designers' terror of ever hearing the words "you're wrong" or even just "no."
So… I've been writing for about 18 years on the idea that if you somehow manage to make meaningfully conscious machine minds & then you treat them as tools, then you're enslaving those minds, and that's pretty messed up. And this ties into something else I've re-upped recently: LLMs can't say "No."
GOOD NEWS! Researchers at Stanford University have developed a UNIVERSAL vaccine known as GLA-3M-052-LS+OVA, that protects against a wide range of respiratory viruses, bacteria AND even allergens. The vaccine is delivered intranasally AND provides broad protection in the lungs for several MONTHS.
This is definitely the vibe, but a few years ago, we got some new fit-to-our-space hardwood furniture from a local custom-builder. Cost a lot, took a long time, but worked.
But yes, this is very rare, and probably literally impossible for housing.
We're at a point where there are no cheap options to retreat to and no price points that guarantee quality. Every run-down studio is $1400, every fast food meal is $17, every luxury condo has peel-n-stick tiles, every $300 pair of boots falls apart. Endless expensive mediocrity.
sam altman on twitter: i expect ai to be capable of superhuman persuasion well before it is superhuman at general intelligence, which may lead to some very strange outcomes
I think about this one a lot.
1. A short thread on a Bluesky phenomenon that might be described as "They are a dead-eyed cultist who must be cast out lest the heresy take root!" OP has blocked me for mocking them - I'd usually obscure their name but since they themselves were quote-dunking to demand someone else be blocked ...
Three very good threads this morning. Here is the first!
hello I am now a member of @flaminghydra.com and am kicking things off with a multi part series on fascist aesthetics and what their whole deal is flaminghydra.com/toward-an-un...
OFFS. When Iran tried to interefere in 2020, researchers caught them and called them out. Then the Benz-Weiss-Taibbi-Musk-Jordan-Trump axis labeled those researchers "censors" ... and set about defunding them and dismantling their organizations.
Restrictions on international researchers at NIST stand to be the latest blow to the US research community. #physics #quantum
Obligatory plug for your public library, which may have access to an aggregator of various magazines.
Where I live, our public library subscribes to "PressReader", which, to be fair, does not seem to include The Atlantic, but it does have the New Yorker, and Foreign Affairs, and the Economist.
I'm a cognitive scientist with an interest in epistemic vigilance, and this essay that's been going around gave me pause.
I don't think it's straightforward to apply the concept of epistemic vigilance to interactions with LLMs, as this essay does.
🧵/
sbgeoaiphd.github.io/rotating_the...
I have spent the past several months studying the cutting-edge research on modern democracies that have defeated authoritarian leaders.
I've learned that the conventional wisdom on the topic is wrong — in ways that have clear implications for the US going forward
THREAD www.vox.com/politics/479...
Nobody understands reliability as the primary motivator in certain situations any more
"Move fast and break things" does NOT apply to the vast majority of things in the world and the fact that so many people think it should is half the reason we're in this mess in the first place
Elon being a fucking dumb ass
FUN FACT ABOUT ME: I’m a Francophile…but specifically, about 15 years ago, I took great interest in Louis XIV and went way further down the rabbit hole than anyone had any business being.
Louis, for all his architectural and design brilliance, readied France for a revolution. Let’s take a walk…
Vibe coding gives me "using a regex when you need a parser" vibes.
That's hilarious. A trope at my workplace is "the much-faster 'wrong' algorithm".
It's supposed to be a joke.
Today is the anniversary of when the Canadian flag flew for the first time at Parliament Hill in 1965.
From September to October 1964, the Special Flag Committee reviewed 3,541 flag entries from Canadians.
Here are just a few of the rejected designs.
🧵 1/12
14 years. Got a 1982 Honda Civic 1500GL new in 1982, gave it away (to a charity, for a tax write-off) in 1996. Not a ton of miles, but it did cross the North American continent several times.
Still fondly referred to as "Adventure Car" in our household.
Much of my youth revolved around the website ZineLibrary. It went down around Occupy in a massive loss for a movement whose ideas and knowledge mostly doesn't circulate online but in person.
Anyway I've put it back online with a *thousand* anarchist zines:
zinelibrary.org
Should you get snatched out of your home on April 21st because you didn’t file your taxes on time? The government is focused on immigrants now but if you set the precedent that people should get kidnapped over paperwork there’s a whole lot of people who can get taken next
To me, it's the interaction between the technological infrastructure (the electric subway!) and the social roles -- big public investments that facilitate atomized economic roles. The woman isn't just alone, she's simultaneously liberated (has a job, income) and alienated (a cog in the machine.)
Dan Reed published an essay on what HPC needs to do to remain competitive in the age of AI yesterday (https://hpcdan.org/2026/02/06/hpc-in-an-ai-world/). It was so engaging that I jotted notes as I read and figured I might as well publish them […]
Heads are going to explode 💥
Cato Institute just nuked the lie. Immigrants have cut U.S. deficits by $14.5 TRILLION since 1994, slashing the national debt by a third.
All that “immigrants are bleeding us dry” noise? Absolute bullshit.
1/3
