Hacker News - It's not for hackers and it's hardly news.

So, I have some more Lobsters drama this time. I still haven't started watching his stream, since I can't get it to work, but reading the transcripts should be good enough. I was tracing this article through the moderation log, since I lost track of it, and noticed both wired.com and nytimes.com have been banned, interestingly. Anyway, here's the story which is my focus:
https://lobste.rs/s/4onklj/ai_training_shouldn_t_erase_authorship (archive)

Here's the article serving as the topic:
https://justine.lol/history/ (archive)

The article's not unreasonable for the most part, but I don't really care. User WilhelmVonWeiner reports the author as a wrongthinker:
AI Training Shouldn't Erase Authorship  Lobsters.png
User friendlysock has a reasonable opinion:
AI Training Shouldn't Erase Authorship  Lobsters 2.png
WilhelmVonWeiner said:
Yeah, it’s about the badthink wrongtribe brush. It’s not about indicating to people “maybe people who talk like this aren’t good people”, or at least just raising awareness that popular software engineer Justine Tunney is not a good person with good intentions. If one person googles “Justine Tunney slavery” I’ve elevated the conversation to a personally satisfactory degree.
I don't currently have access to a convenient screenshotting tool, so multiple images will have to do:
AI Training Shouldn't Erase Authorship  Lobsters 4.png
Hector Martin's tranny alter ego made the HN first page
I found that conversation to be interesting. It's funny to watch these fools focus on a very narrow form of correctness. Preconditions and postconditions would be key to proving parts of the Linux kernel correct, while also documenting their behaviour, but that'll never get done. This resistance to the cult amused me a little: Asahi Lina (朝日リナ)  nullptrlive I think people really don't ap… - VT Social 4.pngAsahi Lina (朝日リナ)  nullptrlive I think people really don't ap… - VT Social 3.pngAsahi Lina (朝日リナ)  nullptrlive I think people really don't ap… - VT Social 2.png
Ada status: mentioned.
Asahi Lina (朝日リナ)  nullptrlive I think people really don't ap… - VT Social.png
This was also interesting on Lobsters:
https://lobste.rs/s/mf7guc/leak_facebook_partner_brags_about (archive)
In Leak, Facebook Partner Brags About Listening to Your Phone’s Microphone to Serve Ads for St...png
You claim to have been raped by a pack of niggers, and yet there were only two niggers. Curious.

I'll conclude this post with one final funny thing I noticed:
https://news.ycombinator.com/item?id=41356100#41376171 (archive)
When A.I.'s Output Is a Threat to A.I. Itself  Hacker News.png
Just be careful that the entire training set doesn't have the moral values of 100 years ago!
Yes, throw every historical document into a blender, but make absolutely, positively certain the resulting amalgamation doesn't express wrongthink! Expressed this way, it reads like something out of a science-fiction horror story, now doesn't it?
 
Yes, throw every historical document into a blender, but make absolutely, positively certain the resulting amalgamation doesn't express wrongthink! Expressed this way, it reads like something out of a science-fiction horror story, now doesn't it?
Best start believing in sci fi horror stories
1725912191245.png
 
a.png
b.png
Earlier today someone hacked a prominent Meta engineer's HN account and posted a link to a fake blog post saying he was resigning because of a "backdoor built into Meta’s front-end infrastructure".
Of course in true HN fashion you aren't allowed to delete posts, so the posts from the hacker still remain up on his profile.
Why I'm Leaving Meta [A]
HN Thread [A]
Screen Shot 2024-09-21 at 20.34.01-fullpage.png
 
Last edited:
View attachment 6441803
Someone earlier today hacked a prominent Meta engineer's HN profile and posted a link to a fake blog post saying he was resigning because of a "backdoor built into Meta’s front-end infrastructure".
Of course in true HN fashion you aren't allowed to delete posts, so the posts from the hacker still remain up on his profile.
Why I'm Leaving Meta [A]
HN Thread [A]
That's a nice find. The first link is malformed.
 
This thread provides a look into the psychology of the Hacker News user, an entire thread of authoritarians demanding that daddy government protect consenting adults from making their own decisions with their own money, a distant culture from a site that supposedly believes in risk taking and startups but is now just a site for aging bay area bigtech jeets and troons to be pro-regulation, pro-censorship, and pro-monopoly.

 
This thread provides a look into the psychology of the Hacker News user, an entire thread of authoritarians demanding that daddy government protect consenting adults from making their own decisions with their own money, a distant culture from a site that supposedly believes in risk taking and startups but is now just a site for aging bay area bigtech jeets and troons to be pro-regulation, pro-censorship, and pro-monopoly.

I was curious if they were going to defend weed zombies. And, of course they are.
2024-09-28_10-08.png
 
Hacker Nudes discusses how AI is shitting up the internet: https://news.ycombinator.com/item?id=41667652 (https://archive.ph/reRtu)
Somehow John Nagle (animats) is the voice of reason as usual.

Some personal favorites:
hn-ai-092824.jpg

bschmidt3 is a personal lolcow that I have marked to watch. I can't tell this this person is being facetious or is actually fucking deranged. Clearly if you don't like LLMs shitting up the entire internet you're a luddite spreading FUD for extremely evil megacorps and you hate freedom and aren't a hacker.
hn-ai-2-092824.jpg

Skynet is literally going to be up our asses tomorrow because an LLM produced some text that appeared half-way coherent.

What is at least somewhat interesting is that the sentiment on AI outside of obvious cryptobro types seems to be souring, which feels like a good indication that expectations are coming back down to earth. There will probably be useful things to come out of LLMs (the work around less hyped adjacent technology like semantic search is big for me) and shit like slop NPC dialogues and pulp lit shit, but it's probably not going to displace anything of any real use. I think we're close to seeing a lot of salt flowing when the VCs and investors realize they've been bagholding what amounts to the latest crypto rugpull.
 
Hacker Nudes discusses how AI is shitting up the internet: https://news.ycombinator.com/item?id=41667652 (https://archive.ph/reRtu)
Somehow John Nagle (animats) is the voice of reason as usual.

Some personal favorites:
View attachment 6464829
bschmidt3 is a personal lolcow that I have marked to watch. I can't tell this this person is being facetious or is actually fucking deranged. Clearly if you don't like LLMs shitting up the entire internet you're a luddite spreading FUD for extremely evil megacorps and you hate freedom and aren't a hacker.
View attachment 6464842
Skynet is literally going to be up our asses tomorrow because an LLM produced some text that appeared half-way coherent.

What is at least somewhat interesting is that the sentiment on AI outside of obvious cryptobro types seems to be souring, which feels like a good indication that expectations are coming back down to earth. There will probably be useful things to come out of LLMs (the work around less hyped adjacent technology like semantic search is big for me) and shit like slop NPC dialogues and pulp lit shit, but it's probably not going to displace anything of any real use. I think we're close to seeing a lot of salt flowing when the VCs and investors realize they've been bagholding what amounts to the latest crypto rugpull.

HN would never admit it but the AI rugpull is going to be worse than crypto rugpulls.

Sure in crypto many tokens are worthless scams, VC premine garbage, and NFTs got rugged completely but the underlying businesses of the exchanges were profitable from day 1. CZ is the richest Canadian for a reason. There will always be a use for a centralized service that matches buyers with sellers and degenerate market makers will always want to participate to make money on extremely volatile markets where traders use 100x leverage.

But monetizing AI seems completely doomed from the outset. There's a reason all this is subsized by the richest tech companies.
Every business that tries to build a wrapper off of an API by a big company? Big company shuts it down and makes their own version. Meta is releasing models to destroy OpenAIs business prospects. People will want uncucked AIs and they will have to use non-American models because America will probably cuck out models for "misinformation" or some bullshit reason.

This guy basically says the numbers don't work.
 
This thread provides a look into the psychology of the Hacker News user, an entire thread of authoritarians demanding that daddy government protect consenting adults from making their own decisions with their own money, a distant culture from a site that supposedly believes in risk taking and startups but is now just a site for aging bay area bigtech jeets and troons to be pro-regulation, pro-censorship, and pro-monopoly.
When you grow up you're gonna realize that there's a lot of "consenting adults" getting into mindboggling debt because of gambling and that this affects you directly by impoverishing entire regions, increasing crime, and making quasi-criminal enterprises richer. I sure as hell do want "daddy government" to prevent people from spending money on pointless predatory shit like this.

Somebody mentioned in that thread that the poorest Brazilians spend something like $450M a month on sports gambling. 20% of their welfare checks. That's money that could be spent on infrastructure, business, raising quality of life, and getting out of poverty. Instead it's just flushed down the john and laundering drug money. If you're defending this, you're a gibbering pants-on-head retard.
 
Some jeets stole code from another YC company in order to get into YC, and orange reddit HN flings shit at each other about it in the comments -

https://news.ycombinator.com/item?id=41697032

Archived in case dang deep-sixes it, in the same thread an HN user claims that nepotism routinely goes on behind the scenes in picking who gets into YC cohorts, and that was the main reason PearAI even got picked at all -

https://archive.is/tAhhm
 
Some jeets stole code from another YC company in order to get into YC, and orange reddit HN flings shit at each other about it in the comments -

https://news.ycombinator.com/item?id=41697032

Archived in case dang deep-sixes it, in the same thread an HN user claims that nepotism routinely goes on behind the scenes in picking who gets into YC cohorts, and that was the main reason PearAI even got picked at all -

https://archive.is/tAhhm

All VCs are like this re:nepotism and other skeezy behaviors (having been in that scene for a while), but YC are just extra extra smug and insufferable. As far as stolen code, between this sort of shit, AI being "trained" on everything regardless of licenses (which just amounts to license laundering TBH), and jeets and other undesirables just generally stealing all sorts of shit, I don't feel like there's a lot of incentive to open source things anymore, especially if it has any chance of being commercially useful (unless of course someone is paying you to do it).

It's been hilarious to watch all of the naive retards over the last 15+ years talk about how their bullshit is "going to change the world" and "how they have a mission to do good" when their investors do not give a fat fuck about anyone other than themselves, and their ideas basically involve skirting or breaking the law or otherwise just lighting money on fire to benefit other members of the laptop class. YC is probably some of the worst about that sort of hypocrisy, and there's a lot of 20 somethings that just eat that shit up hook line and sinker.
 
I'm not sure if this is the right thread, but a poster who really annoys me is withoutboats - https://without.boats/

He's a Rust guy who contributed to some RFCs, and writes 20,000 word blog posts about very deep rust internals. He definitely has some technical chops, but he does an awful lot of yapping, and seemingly not much else...

2024-10-21-152828_1292x668_scrot.png

Reason I'm in this thread is that he posts arrogant and vague comments on HN, and when people disagree, he simply points to his massive body of writing and says "read it"

2024-10-21-152647_1142x553_scrot.png

(link, a)

What got my goat was in the comments on a recent article discussing ways that Rust can be improved/matured, he basically dismissed all suggestions as being impossible, and called the author's writing "sophomoric," again telling him to go read his 20,000 word essays.

2024-10-21-154047_1155x267_scrot.png

(link, a)

Also, his username is a Foucault quote...

2024-10-21-154444_607x121_scrot.png

It's a lot of little things, but sometimes certain people just wig me out.
 
Back