I used to always consider it unrealistic in sci-fi/fantasy stories when there was somehow an intelligence that could be talked to that had no/few personal experiences but somehow had a lot of general world knowledge. But now that's exactly what real LLMs are like.

This is honestly one of the most shocking images coming out of this administration and that’s saying something

Murshed Zaheed@murshedz.bsky.social · last wk.

“Trump has rejected disaster aid for Democratic-run states at the highest rate in the 47-year history of the Federal Emergency Management Agency.” Right now people in blue states are basically paying federal taxes as if they’re colonized territories of red states. www.politico.com/news/2026/03...

someone needs to make a game engine named "Godoh" just to add confusion in every discussion about how "Godot" is pronounced

(deltarune posting) inventing a new form of mental illness by taking this idea too far and jokingly telling claude to beam "Proceed" into my neuralink at appropriate times

JuanMa 🇨🇴@jgirco.bsky.social · 2w ago

Maybe what I need in my life is a voice that every morning tells me "You should do a breakthrough" and one that periodically whispers "Please continue".

I wrote about the completely wild incident where OpenAI were testing a new model and it broke out of its sandbox and broke INTO Hugging Face to steal the answers to the benchmark simonwillison.net/2026/Jul/22/...

every time I've fixed an exploitable memory safety vulnerability in C/C++, I do it by noticing the code's violations of Rust's borrowing rules and then I translate the idiomatic Rust solution back to C/C++.

fable had to return to her home planet

i'm pro technology, but doing it right, which means as cleanly and safely as practical, and i refuse to accept the richest companies on earth can't afford basic environmental protections.

post malone ergo propter malone@proptermalone.bsky.social · 3mo ago

if you show me a negative externality I am basically always going to say “yeah they should be made to pay for that”

Maxine 💃🏼@maxine.science · 3mo ago

An external write-up direct from the raw chain of thought: maxine.science/erdos-unit-d...

Everyone using dependency managers like npm, pypi, cargo, etc, should configure it to ignore packages that are less than a day or two old. Really it should be the default. None of the recent major compromised packages had malicious versions up that long. cooldowns.dev#javascript-e...

Feross@feross.bsky.social · 3mo ago

🚨 Active supply chain attacks on both PyPI and npm today. lightning 2.6.2/2.6.3 and intercom-client 7.0.4 are compromised with credential-stealing malware. Same attacker, same tooling, two ecosystems. Downgrade immediately.

Yeah some unsolicated hot takes in no particular order 1. React is better than everything that came before it and I haven't seen anything do it better 2. Hooks are genius 3. If you've never done frontend on a moderately large app, you probably underestimate frontend engineering

dan@danabra.mov · 4mo ago

it’s always funny to me when people try to paint react as some kind of “big company” propaganda. it was like one person’s wild idea and then there was at most ~5 people maintaining it for years. people HATED it when fb open sourced it. the only reason it won is because the model was actually good

I've been curious to see how new sci-fi deals with the current AI boom. So far I've just seen it done extremely eyeroll-inducingly: "Our setting's AI has nothing to do with the *fake* unethical AI of the past! (Our AI is an evil guy who kills people.)" (Marathon, High on Life 2)

they are putting my brain inside the computer tomorrow

for what it's worth, the vulnerabilities described in the Mythos article that it found and exploited were of usual kinds. There are no mysterious move 37s. If AI discovers a new class of vulnerability that affects code patterns we strongly thought were safe then I'll panic.

When ChatGPT was new and everyone was anxious to understand the specific pace of AI progress, I figured a significant milestone to watch for was when it could find and exploit brand new software vulnerabilities on command. It's slightly unnerving to see the milestone met today.

Why@why.bsky.world · 4mo ago

New Claude found zero-day's in OpenBSD, ffmpeg, Linux and FreeBSD.

"Over 99% of the vulnerabilities we’ve found have not yet been patched"... "Mythos Preview is capable of identifying and then exploiting zero-day vulnerabilities in every major operating system and every major web browser when directed by a user to do so ... oldest ... 27-year-old bug in OpenBSD"

Alex Becker@alexcbecker.net · 4mo ago

This is the real deal folks red.anthropic.com/2026/mythos-...

The post-quantum cryptography timeline accelerating at the same time that mythos is creating mass CVEs is really threatening to alter my roadmaps

kira is now better at posting than i am, apparently i am losing my job to automation

kira ☆@kira.ws · 4mo ago

people with robot girl pfps posting 'datacenters are destroying the planet' from their cyberpunk aesthetic accounts. bestie my culture is not your costume

me reading the replies to the @attie.ai announcement

Matt Keeter@mattkeeter.com · 10mo ago

Bluesky's moderation challenges stem from the fact that it was founded by Blacklaws, but the average user is a Cousin. In this essay I will

I wonder how much Sora shutting down has to do with Iran using AI for propaganda videos lately. Kind of awkward for a business to be visibly associated with that.

"To achieve this, we implemented a WebAssembly interpreter inside the transformer weights." sentences with terrifying aura

Eugene Vinitsky 🍒@eugenevinitsky.bsky.social · 5mo ago

www.percepta.ai/blog/can-llm... As a research lark at Percepta, Christos embedded a computer into an LLM, showed that it could solve the hardest Sudokus, and then as a side bonus built an exponentially faster attention

and gwern.net, of course

Macil@macil.tech · 5mo ago

It's easy to forget that the average person's foray into philosophy consists not of completing the Sequences but reading just one or two of them

Permutation City (Egan, '94)

Coffee Indiana@coffeeindiana.bsky.social · 5mo ago

If Claude is conscious then wouldn't math be conscious as well?

aside from the AI angle, what makes the Anthropic case important is that it is -- unless there is something I am forgetting -- the first time a major tech company has stood up to Trump instead of accommodating him

ngl. to me it's a miracle that i can talk to a computer like this, and it knows what i'm talking about and does the right fix. many engineers would struggle with this (including myself)

started noticing that it's common for music apps with sleep timers to have the options max out at 1 hour, and it has me questioning why multiple companies would pick a max time that is so obviously lower than the usual time it takes to fall asleep

anyone looking to replace their use of Discord with something more private shouldn't just jump ship to another proprietary system that's run the same way, and should instead look for something open source, end-to-end encrypted, and distributed/federated, like Element(Matrix).

if the Turing test was passed in a way that matters then Moltbook would be indistinguishable or better than other social media. If you don't find Moltbook as interesting, or you find yourself talking about the funny things on Moltbook with humans instead of AIs there, then we haven't reached AGI yet

Ethan Mollick@emollick.bsky.social · 6mo ago

A pretty bold comment in Nature written by linguists, computer scientists and philosophers declaring that AGI has been achieved. "By reasonable standards, including Turing’s own, we have artificial systems that are generally intelligent. The long-standing problem of creating AGI has been solved."

I want off Mr. Epstein's Wild Ride

This case isn't saying that Google devices have an intentional secret recording mode. It's about Google devices being too sensitive in detecting the "Ok Google" activation phrase and triggering the normal voice assistant feature. www.googleassistantprivacylitigation.com/docs/4th%20A...