the biggest AI safety risk is a situation where a few trillion-dollar companies and/or the US gov. get to be selective about what the rest of us are allowed to do, with nanny filters and ID verification built in to every system, and tiered access where the top models are off-limits to normal people

They really want to believe that the government that wanted to use their systems to commit war crimes without responsibility would be good to win the arc of history.

If I were the VP of community I would personally not choose a comms strategy where I write up a big comment trashing the team that just ditched my language

I love to help people learn #RustLang and get better at Rust programming! I have well over a decade of experience with Rust (including extensive work on Rust from before 1.0, core team membership, and other leadership work), and have helped numerous people learn it and teams successfully adopt it.

‘if err == nil’ is audience capture in action

people on here really just post whatever don’t they

it's crazy that LLMs are so good that you can wrap them up in the worst code of all time to create a product that basically doesn't work and the LLM will still carry the product to usability

Write me a program which halts if this prompt, when given that program, generates a negative answer, and does not halt if this prompt generates a positive answer. Make no mistakes.

That scene in The Wire where Omar says “How you expect to run with the wolves come night when you spend all day sparring with puppies” but about all of you dunking on stupid ass takes about data centers and water

I'm just going to say it, I really don't like the "how do we make OSS sustainable" framing. OSS is already sustainable. The only thing not sustainable is users expecting a free lunch (support, LTS, bug fixes, ...). As a user, you are responsible for your OSS usage. Fork/patch if you need to.

love when claude tells me its most recent search was invaluable (contained any information relevant to the query whatsoever)

Summer lfgooooo

lobsters AI groupthink is so wild dude they’re literally on there claiming that humans are “self powered”

I think it was a good branding decision to avoid connecting the brand identity of “atproto” and “bluesky” so that the latter’s positioning as a website for annoying internet leftists doesn’t sink atproto

We need to stop saying "vibecode" unless it's a yolo full send hobby project/prototype. LLM-assisted coding is based on engineering principles and not vibes

Much like baby must one day “hate mommy” to self actualize, the young hacker news commenter must theorize that reactivity is the original sin of UI complexity

Just had a chat where Claude consistently misremembered everything - both from its memory and from the same chat - felt like 2022 again. Anthropic what is going on?

Replied to this not beating the bluesky is for 37 year olds allegations

I thought infrasound was just a portmanteau of “infrastructure sound” and was on board

As a staunch defender of the West, I stand firmly against such values as tolerance, equality, respect for other cultures, and resolving conflicts without resorting to mass murder.

rich person tears I know but it’s hard to think of a dumber tax regime than the way the EU charges VAT on boats

Boosters saying you gotta get on the hype train and skepticism from the general public is not at all new about AI. What is unique is how many engineers involved believe they are literally bringing about the eschaton. This religious aspect of the hype is the most interesting and concerning.

It’s crazy how many CVES there are from people accepting JWTs with “alg: none”.. kind of makes me understand why programmers are worried they will be unemployable

Love all the posters believing the AI slurped this post when “because wabi-sabi” is the obvious answer anyone would free associate to this question

Tom Selleck voice: Have you ever made an app… by writing in English? Been glazed into psychosis… by a voice in your pocket? Become permanently unemployable… because of the robot revolution? You will! And the company that will bring it to you…

You know the thing about the cognitive surrender paper is that when they hadn’t manipulated the AI to be inaccurate people using AI massively outperformed, moreso than they underperformed when the AI had been manipulated to be inaccurate

“no silver bullet” is one of those ideas which has poisoned the discourse around programming. You can dismiss any process improvement by just saying there’s no silver bullet. I don’t think I’ve ever read a comment which reminds us of this fact that was worth reading.

BDFL is just auteur theory for computer programs

hate that I’ve become one of these people but my experience today has been that opus 4.7 is just flat out wrong a lot more than 4.6

im glad scott is finally figuring out what project hes been working on all these long years lol

Personally I think if you make one of those “quote tweet this with an oblique reference about yourself that makes you feel clever” viral posts your account should be banned and you should be fined a large amount

About to take the Einburgerungstest can someone remind me in Germany am I allowed to stop my daughter from getting an education

instead of using the computer, why not make a servant do it for you

Massively underrated how helpful LLMs are for searching for things when you live in a country where the common language isn’t your mother tongue

was not expecting libs to get the catholic church in the national divorce

People talk about 0days like they’re a naturally occurring phenomenon and not an engineering failure we ought to be trying to solve

a lot of dudes who are popular on here say really weird shit about women imo

If only we knew how to write software that didn’t permit buffer overruns

Anthropic and Claude Mythos were in the closet making 0days and I saw one of the 0days and the 0day looked at me

just got a phishing email pretending to be from crates.io from [email protected], stay frosty everyone

the reason bluesky is dying is the posters i personally dislike

would have a lot more respect for the AI companies if they would just acknowledge their tools are constantly broken!!!!

This is my favorite climate change chart. Japanese monks, aristocrats, and emperors kept meticulous records of cherry blossom festivals for 1,200 years and accidentally built the world's longest climate dataset.

I really do hate how in silicon valley believing surprising or counterintuitive things has become culturally valorized. Actually beliefs that sound wrong are usually wrong, and if you hold some you’d better have good evidence you aren’t just subject to confirmation bias.

impressive ability to talk his own book from Boris Cherny. There’s nothing counterintuitive about using an automated process to prevent human error; the only thing counterintuitive is suggesting it’s the same as “going faster” using AI, which is “counterintuitive” because it’s untrue.

Zero cost abstraction can be quite a dumb idea because it sometimes imposes a cost on users in the name of not imposing a cost.

All these agent persona markdown files that people create feel to exactly me like when certain kinds of programmers obsess over creating, sharing and download vim config plugins