Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.
This is not debate club. Unless it’s amusing debate.
For actually-good tech, you want our NotAwfulTech community
so a couple months ago it was noted that mattermost’s website had gone very problematic, extreme emphasis on pitching to mil/int with very little indicating its foss roots/nature
and now it seems they’re going for rent extraction even on the self-hosted
I have a webserver using Ubuntu. Are any of the popular LLM-poisoning solutions packaged for that?
I ran across this today and while I had fun mashing together the Sequences, Moby Dick and 1984, a pre-packaged solution would be even better.
Dont know the exact details, but there now is a fork from the popular doom engine gzDoom called, UZDoom, because the initial dev came out retirement out of nowhere and started to vibe code, causing all the others to quit.
More here https://zdoom.org/news#p80996
you know, I shouldn't be surprised by the extremely toxic lead developer to prompt enthusiast pipeline, but... slopcode in gzDoom of all things? fucking why?
I heard there was even more wrong with what he did than was described in the quite nicely written zdoom post (who clearly just want to move past the drama) but that was third hand information. Altogether quite odd behavior of the old dev. Clearly he didn't learn from Frozen.
Via a prev sneer, here's a Bloomberg post about call center workers being accused of being LLMs:
I used to work in a call-center adjacent industry and the amount of crap some employers used to make sure their workers sounded as much as robots as possible was astounding.
The FSF considers large language models, failing to recognise the purpose of plagiarism machines in the process.
Comments are also pretty sneerable - main thread's a bunch of promptfondlers trying to gas up spicy autocomplete, but I did find some guy claiming programmers steal and plagiarise much like LLMs do.
Unless you’ve only ever used ChatGPT, you will know that LLM-produced code is not the result of a single prompt, not even a conversation, but rather a workflow that often goes as such:
- Discuss a problem with the LLM. The LLM autonomously reads large parts of the repository you’re working in, during the discussion.
- Ask it to write a plan. Edit the plan. Ask it about the edited plan. Edit it some more.
- Repeatedly restart the LLM, asking it to code different parts of the plan. Debug the results. Write some code yourself. Create, rebase, or otherwise play around with the repository; keep multiple branches of potential code.
- Go back and edit the original plan, now that you know what might work. Port some unit tests back in time, sometimes.
- Repeat until done.
is so stupid
rubber duck debugging with a cost multiplier magnitude coming in at a couple of zeroes tail end of the figure
and worse results
it sounds like you've only used the llm with the anus logo
have you tried this other llm^[https://velvetshark.com/ai-company-logos-that-look-like-buttholes]?
image description
picture of various tech company's logos arranged in the outline of a butt with claude's logo in the center
claude's logo looks like an anus
noticed someone like "comments are lazy! never use them! write better code!"
A:
I once had someone tell me to my face that comments were a code smell.
Depends on how they are used. They are extremely useful when they add context that is otherwise not present in the code, but way too often I see people simply restating what the code does in the comment. That's not very useful and can become confusing if the code later gets updated without updating the associated comment.
this seems counterintuitive but... comments are the best, name of the function but longer are the worst. Plain text summary of a huge chunk of code that I really should have taken the time to break up instead of writing a novella about it are somewhere in the middle.
I feel a lot of bad comment practices are downstream of javascript relying on jsdoc to act like a real language.
I've found that people who talk about "code smell" generally should not be listened to as it's entirely vibes-based-on-the-last-medium-post-i-read-this-morning. I had a dipshit manager tell me that he didn't like my use of decorators (in python mind you) because it was a "code smell" and recommended I read "clean" code, and I immediately threw every other opinion he had in the trash.
Managers gonna manage, but having a term for bad code that works that is more palatable than 'amateur hour' isn't inherently bad imo.
Worst i've heard is some company forbidding LINQ in C#, which in python terms is forcing you to always use for-loops in place of filter/map/reduce and comprehensions and other stuff like pandas.groupby
that's truly assenine as LINQ is supposed to be one of the few reasons to want to use C#
So anybody else noticing regular outages here today or was it just me?
uggggh yep sorry, I meant to deploy some changes that would mitigate this a couple nights ago but had a bunch of things crop up. I’ll do my best to work it in tonight!
some specifics for the changes to expect:
Nah it happens, at least it isnt on my end.
current issue is likely the most recent bout scraper waves (based on it matching data myself and others have seen elsewhere); suggestions have been made, so hopefully soon(tm) changes
(don’t mean to speak on @self’s behalf ito plans, mind. that’s for them to do :D)
nah same for me
Yeah, there's been a bunch of outages recently.
Guys, it's a problem that GenPop thinks "AI" is useless, because that means they will deny the inevitable coming existence of the Evil Robot God and make it harder to stop it
https://www.lesswrong.com/posts/W2dTrfTsGtFiwG5hM/origins-and-dangers-of-future-ai-capability-denial
The whole thing has to be read to be believed. But there are some fun nuggets:
I can imagine a world where AI becomes extremely good at flipping burgers or driving cars well before it learns how to write software or […] protein folding or playing board games
That’s because despite Moravec’s paradox being noted down in the 80s (y’know, the last ai winter) there’s still a certain kind of asshole who thinks that flipping burgers is an easy task performed by stupid people but playing go is somehow the height of human intellect, despite 40+ years of evidence to the contrary.
i'm only at the beginning, but this already stuck out to me:
It would be extremely bad if many people come to believe this: arguments about existential risk mostly rely on the assumption that AI is capable, so they fall flat for people who don't agree with that. I think we should be emphasizing the core capability of AI more and talking about x-risks less.
"Guys we need to pump this ~~bubble~~ Rational Stock Appreciation Trend more, the plebs are starting to think it's all bullshit!!"
Well, it is true that computer programs have far surpassed humans in board games. They are very well suited for it. It just has nothing to do with the hypothesized abilities of future "AI" as rationalists conceive them.
In lighter news, this anti-LLM rhyme made me chuckle:
I will not talk with a chatbot
I do not want it while I shopI do not want it on Windows X-box
I do not want it in FirefoxI do not want it in my house
I do not want it on my mouse
I do not want it here or there
I do not want it anywhere.I do not want AI and Spam
I do not want them Sam-Alt-Man
For something lighter, here's an AI bro getting wowed by the shittiest "video game" I've ever seen (trust me, the screenshot doesn't do it justice):
In lieu of sneering this shit, I'd like to argue that arts education should become mandatory for all students post-bubble, regardless of their profession. STEM, humanities, tech, doesn't matter - give them four years of art so they don't turn out like this guy.
Amazon unveils AI-Powered AR glasses for delivery drivers.
The response I've seen from actual drivers has been somewhere between "eh? I guess?" and "fuck you, pay me".
Does it have GPS so it can flash "WRONG HOUSE, FUCKWIT"
JFC some CSAM collector stormed a Wikipedia conference with a gun https://slate.com/technology/2025/10/wikipedia-editors-conference-gunman-culture-war.html