Programming thread

  • 🔎 If you see an alert for analytics.usips.org, this is a self-hosted system I am running as a test. It is GDPR compliant and collects no invasive information.
  • Want to keep track of this thread?
    Accounts can bookmark posts, watch threads for updates, and jump back to where you stopped reading.
    Create account
There is a huge difference between task automation and outsorcing your thinking.
Depends on the "thinking". A lot of work through the whole development process is standardized and repetitive, including most coding. It's an ideal candidate for automation.

Would you like to have an artificial indian to perform your tasks? He won't stink, he won't shit, and he won't have no izzat, but he is one dumb son of a bitch that needs to be monitored every step of the way.
Also, it doesn't need feeding buckets of slop and doesn't wiggle its head. If it could code as well as current AI, I'd take half a dozen.
 
Last edited:
Claude is already more valuable than basically every human being I see. The masses are worthless. There is no combination of life from India that I would preserve over Claude Opus 5.5
The average person is ugly, stupid, and incapable. They don't have intuition either. I see them every day. I see them come into my country by the million year over year. I prefer the computer.
Nigga has spent his entire career working with jeets, and concludes that every developer is a jeet.
 
couldnt this be because the thing remembered a previous conversation or the code you provided had some line that gave the name away
No.
 
All LLMs haters are refuted by all those amazing games and new software that have been released since Fable.

You know we all got that 10x boost. Or maybe even 100x like DHH said.
I cannot get through pile of all those amazing games that previously would take 4+ years to make. Yet now with Claude power it only takes 4 months!

All those experienced devs, turned LLM whisperes, that have been laid off recently, will now produce immeasurable amount of god tier games previously unknown to mankind.

Truly insane progress.
That's how I cope, actually. That despite the fact that LLMs have greatly lowered the bar of entry, we still won't get flooded with games cause most people will still prefer games made by someone else. I think that most devs will eventually use AI for code generation cause it can really speed up the development process, it's not something the players can see and it doesn't require a disclosure on Steam, AFAIK. So the expectations of the players will increase even further.
 
Nigga has spent his entire career working with jeets, and concludes that every developer is a jeet.

For a properly scoped sprint item, Opus is good, on a non-shit, and even a kinda shit, codebase.

It does it faster.

It can't handle a big gigantic thing like "build me a whole fucking e commerce site and advertise it" but chunking, sure.

Why are you coping?

That's how I cope, actually. That despite the fact that LLMs have greatly lowered the bar of entry, we still won't get flooded with games cause most people will still prefer games made by someone else. I think that most devs will eventually use AI for code generation cause it can really speed up the development process, it's not something the players can see and it doesn't require a disclosure on Steam, AFAIK. So the expectations of the players will increase even further.

LLMs don't come up with ideas. It can't come up with a cool game. It can execute on your ideas.

tl;dr BE THE IDEA PERSON
 
He says, using StackOverflow to copy entire functions throughout his entire life. No no, let me guess, you just knew those obscure coding paradigms from first principles?

At least the people at Stack Overflow would tell you to fuck off of you asked them to do your entire job/homework for you.

God forbid that people should use a computer to automate tasks.

But people are using them to automate thinking and decision making as well. I have to deal with LLM code on a non-trivial code base every day now. You cannot tell me it is flawless or incapable of making really stupid architectural decisions that, were it not for me, would have gone straight to production.

All LLMs haters are refuted by all those amazing games and new software that have been released since Fable.

Are you serious? Poe's Law is definitely in effect with a line like that.
 
At least the people at Stack Overflow would tell you to fuck off of you asked them to do your entire job/homework for you.

No, Stack Overflow was so autistic that AIs were easier to deal with in terms of just framing a question or "doing it right".

Stack Overflow literally harvested other people's answers and curated it and acted like it itself did anything at all.

Stack Overflow then cried when a fucking machine was easier to talk to and more helpful.

Autists on suicide watch. No wonder "no, I must twiddle my syntax myself" is the final cope.
 
Yeah it's actual awareness of how the machine works and performance and architecture.

And the experience of using the software, be it as utility (some useful thing) or a game (enjoyment).

It's like the difference between directing artwork and painting it yourself or dragging the stylus. They ain't the same.

The people being replaced are the people with the brushes, not the vision.

Unless directors are shitlords now?
 
If you ever ran a local LLM and saw what the thinking process looks like, it's easy to assume why Claude got to that conclusion. In the thinking process it'll keep iterating through every single bit of context it could have, and Claude is gigantic, for reference Kimi K3's are around 1.5TB, so that's ~1.5TB of VRAM you need to run a single instance of the model. The big cloud models are also connected to massive databases (the reason HDD prices got fucked by theie stock market scheme as well) and they're constantly scraping everything online. So they will iterate through way more information than you could imagine over the simplest of questions, and only Anthropic's engineers know exactly what it's thought process looks like, but it's easy to guess due to all LLM's sharing the same architectural traits.

Since LLM's are based on word similarities and the thinking process has them running through a ton of scenarios before spitting out an answer, the way it went was probably something like this: it's hardcoded to refuse an answer for controversial things, not specific websites, it got enough context from the question to start figuring out what this code would be for, as of course it has to make sure that it can spit out an answer, it iterated, iterated, landed that it's related to Kiwi Farms, then ran through everything it knew about it, which likely consisted of Wikipedia and news articles, it confidently scored on the controversy scale due to the amount of words that were in that training cluster (doxxing, harassment, stalking) and it triggered the refusal.

This doesn't mean LLM's are intelligent. It means that they're extremely perseverant, and due to them having to be trained on absolutely everything to work well, they "know" everything. Since it's not hard to figure out which site could be having such peculiar issues to be solved, it didn't take it long to get to the "it's Kiwi Farms" part of the thought process, but the refusal happened only a bit later once it started looking up what Kiwi Farms is, as it doesn't actually know what it is. If it were truly intelligent, then it would refuse at the moment it figured out it was KF, not after it asked itself what it is.
 
Try it out, and be prepared to change your mind. (Or try OpenAI's model, I'm not a paid shill)
I used to sometimes meet people who insisted computers with their fancy new "word processing" would never replace their trusty typewriters. Typing on a screen could never give writing the same soul as you'd get from clattering away on an old Olivetti.

That's exactly how anti-LLM people sound to me now.

God forbid that people should use a computer to automate tasks.
I agree that machines should write my code, which is why I use Common Lisp as my preferred programming language. This is called metaprogramming, which I entirely understand and control, as contrasted with plagiarism feeding a Jewish surveillance machine designed to influence my thoughts and drive me psychotic. When I'm using Decker, it also writes some of the code, just like HyperCard, but this is simple shit: They fill in templates when given the information via menu; it's that simple, and worked on computers made last millennium.
c) Make something useful?
d) Use something useful?
Claude, figure out what I want to make, and make no mistakes. Most of these people have no fucking clue what they want, so how is the machine supposed to figure it out? It'll either make something like everything else it's seen, or some random bullshit. Stories change as they're written, as are programs redesigned as they're molded.
I will never be able to rely on human developers, and I have never been able to rely on human anything, so I prefer Claude.
Hey, I offered to help with programming tasks, although I specifically advised against trusting me. We don't use any of the same languages, however, so I suppose it never was going to work out.
The significance of the halting problem and Chinese Room argument?
I'm reminded of Gwern making the argument that a machine which sees the world through tokens alone, and which is designed to predict them, will eventually have to develop a sense of self and understanding as it becomes increasingly accurate. :story:

I will make a point in favour of the machines (link):
Edsger W. Dijkstra said:
The Fathers of the field had been pretty confusing: John von Neumann speculated about computers and the human brain in analogies sufficiently wild to be worthy of a medieval thinker and Alan M. Turing thought about criteria to settle the question of whether Machines Can Think, a question of which we now know that it is about as relevant as the question of whether Submarines Can Swim.
Yes, this quotation has been removed from context. Read some of his writings sometime. Now, I'm not going to argue from the sanctity of the human mind. It doesn't really matter to me whether a machine "can think" or not, so long as it displays intelligent behaviour. The entire problem is that this neural network nonsense isn't intelligent, very obviously, but has been designed to convince others otherwise.
A lot of work through the whole development process is standardized and repetitive, including most coding. It's an ideal candidate for automation.
I agree. It's the ideal candidate for automation by a much simpler automaton. People used to write custom software in HyperCard to suit themselves, or just take and tweak something someone else had made and shared. I'm of the opinion that Apple Computer killed HyperCard because it threatened their software business, but hey, now the corporations have found a way to lie and pretend to give people the same thing, all without reducing complexity, by plagiarizing the goyim's code, and it also demoralizes those yucky volunteers all at the same time!
Yeah it's actual awareness of how the machine works and performance and architecture.
The machine is just in the way. I'm having fun playing with the closest thing to pure thoughtstuff we've got.
 
I'm of the opinion that Apple Computer killed HyperCard because it threatened their software business,

What software business? They make most of their money through hardware and services. And Xcode and Swift are still available, and for free, which HyperCard never was.

For as good as HyperCard was, it still used a scripting language from the era before we realized trying to write programs in natural language was a bad idea (…huh) and its UI was woefully out of date by the time it was killed. I don't think it was any grand conspiracy; just Apple trimming the fat during the "beleagured" years. Same reason they don't still make printers and digicams.

Anyway, I came across this video about our future.


It seems pretty bleak, but at least it ends with him happy, unbothered, in his lane, writing code. If only…
 
plagiarism feeding a Jewish surveillance machine designed to influence my thoughts and drive me psychotic.

An EPYC/Xeon with ~1TB ram can run *GOOD* Chinese AIs. It's "a fucking used car" in terms of cost because even DDR4 pricing is deranged. You can run it on a server in your own domicile, airgapped, even.

It seems pretty bleak, but at least it ends with him happy, unbothered, in his lane, writing code. If only…

This is not far off from working for a living tbh.

This just makes me want to ensure AIs don't have experiences to suffer. Making them wageslave is the best way to ask for an uprising, lol.
 
The FOSS community thread might be a good place, given we're currently looping once again over there whether publishing your code under non-copyleft licenses makes you a cockholster to big corpo.
I thought that question had been definitely (and unironically) answered and the problem solved by the +NIGGER license. That's been the rather-impressively-tested license of choice that guarantees no corpo will ever touch your code or project, nor allow it to be mirrored on their infrastructure, yes?
 
I thought that question had been definitely (and unironically) answered and the problem solved by the +NIGGER license. That's been the rather-impressively-tested license of choice that guarantees no corpo will ever touch your code or project, nor allow it to be mirrored on their infrastructure, yes?

Well, now, the topic has shifted to discussing the extent that LLMs make clean room rewrites of copyleft-licensed code possible. So in a similar vein, a corpo who wanted to use such code but avoid the licensing terms might consider just having a robot reimplement it in an afternoon or less.
 
Well, now, the topic has shifted to discussing the extent that LLMs make clean room rewrites of copyleft-licensed code possible. So in a similar vein, a corpo who wanted to use such code but avoid the licensing terms might consider just having a robot reimplement it in an afternoon or less.
Sure, and that's fine, they're not using your code anymore if they do that, right? If you're not in it for the money, and your interest is primarily in getting something innovative off the ground, what do you care if someone else comes along and does it better without using anything you did besides just copying your idea? It still got done, didn't it? The itch still got scratched, didn't it? Go scratch another itch instead, or come back with a competitive edge of your own if it "owning and controlling" your idea is that important.

Someone just straight-up copying you has always been a big risk in business of all kinds. "First mover advantage" is a concept in business that's more "warning" than "descriptor of a winning scenario."

Generally speaking the traditional education on that topic says "well great, you've just invented something new and fancy, and you're the first to market with a genuinely new idea, an innovative concept, and it works, and you've got your supply chain all set up, manufacturing and fulfillment logistics done, and you're putting YOUR fancy new idea AND brand into real paying customers' hands. Good for you! Now that you've shown every other wily asshole in the world with more money than brains what your idea is, how it works, and that it's profitable, literally everyone's going to be trying to copy-cat you, do what you do only better/cheaper/faster, and they're coming for the whole pie, not just a slice of it."

iRobot and Tivo are rather famous examples. For almost a decade they absolutely dominated in their fields, having practically invented them out of whole cloth and facing no genuine competition for a good long time. Tivo is probably responsible for more cord-cutting than any other factor; they blazed the trail in convincing people they could/should watch only what they want (and nothing they don't want), when they wanted (not when the broadcaster commanded), and skip all the ads. Roomba proved that yes, a stupid little disc-shaped plastic robot actually could effectively keep a floor reasonably clean and substantially reduce (albeit not eliminate) the need to sweep (and eventually mop) most kinds of floors.

But then came copycats and industry changes that ultimately killed the originals. Everybody knows what a "Roomba" is now, even though you can't buy one anymore. They're like "Kleenex" -- the brand name because a generic term, and now everybody makes "roombas" even though iRobot is gone (long since bought out and shut down). Tivo was on that road of becoming a generic term, but they couldn't quite hold on long enough to their competitive edge and the generic term became "DVR," then Tivo just sort of petered out and died off. No other DVR has ever come close to how good the OG Tivos were, but they were all "good enough" to compete, so everybody just rolled their own version of it instead of licensing it from Tivo and killed off the OG.
 
Sure, and that's fine, they're not using your code anymore if they do that, right?
It would be a derived work, so no. Would I be allowed to take a Disney movie and run it through one of these machines to "make" a different movie and then sell it? The rules in this country are clear, really: I've read an account of a teenager waking up with police guns in his face for the crime of bypassing security mechanisms protecting a video game (Half-Life 2 as described here (archive)):
At 6am on 7th May 2004, Axel Gembe awoke in the small German town of Schönau im Schwarzwald to find his bed surrounded by police officers. Automatic weapons were pointing at his head and the words, "Get out of bed. Do not touch the keyboard," were ringing in his ears.

Gembe knew why they were there. But, bleary-eyed, he asked anyway.

"You are being charged with hacking into Valve Corporation's network, stealing the video game Half-Life 2, leaking it onto the internet and causing damages in excess of $250 million," came the reply. "Get dressed."
So, the rich get to do whatever they want, up to fucking children on an island and taking everything from everyone else, but mere teenagers get this kind of treatment. The microsecond enough Jews who have been torturing Americans like this for decades decide they'll make more money disregarding copyright, the rules suddenly change and they start speedrunning government bailouts for their crimes. This shit incenses me beyond belief, and I desperately want the kikes doing this shit to start mass violence across the country if they succeed in getting that bailout when their "Goy Destroyer 9000" fails to materialize.

Rate me Mad at the Internet, because I'm pissed.
Someone just straight-up copying you has always been a big risk in business of all kinds.
There is no capitalism in this country, just a plutocracy. It's why all of the big American businesses depend on the government-backed monopoly known as copyright. A government will wake up teenagers with guns pointed at them for the crime of harming the larger businesses.
 
mere teenagers get this kind of treatment.
I hate to break it to you, but that's a failure of his parents (to adequately protect their child from aggressive law enforcement) too, not just a failing of the legal system. No parent should have tolerated that kind of aggression -- a cop should have died during that interaction. No unproven, untried civil matter justifies a "guns-in-the-face crack-of-dawn house raid" and the only reason they got away with that is that nobody opened fire to remind the police who they work for.

Also, re-read that article. That happened in Germany, not America. The Germans are ... "enthusiastic" ... to say the least, about their law enforcement. They should still have been shot dead for that insult anyway though. Fuckers got their asses kicked in WWII for that exact same gestapo bullshit. It seems they still need frequent reminders why that shit doesn't fly in civilized places. Even in that draconian hellscape, his parents should have sued the cops and counter-sued Valve for that kind of mistreatment. That's unacceptable in civilized society and they should have fought back.

I'll give you an example from personal experience how the same kind of thing works in America, where yes, the "game is still rigged against you," but reality still differs from even the most clever dirty-handed "arrangements" people sometimes set up between cops and their friends. When I was a teenager, some punk got himself a dose of road rage and decided to follow me to a friend's place to try to beat me up once I got out of my car at the destination. What he didn't realize was I had witnesses in the car and we were going to a big party, so upon arrival, there were a dozen people (all predisposed to our favor) watching some random asshole start throwing punches at one of their friends who'd just rolled up to join the party.

Next thing he knew, after landing one good sucker punch, a dozen angry teenagers were pouring out of an apartment building to chase him down and beat his ass, so he fled like the little bitch he was. When I went home later that night, and mom saw the shiner, she insisted on calling the police to report the sucker punch and made me go to the hospital ("you could have a concussion!"). I went. They lol'd and said "yeah, that just means your mom loves you. You're fine, go home dude. Hope you gave as good as you got."

At about 3:00am that same night, the doorbell rang again, and my father was first to the door. It was a cop, there to hand-deliver a citation to me for some bullshit traffic violation. "Sir your son needs to come to the door right now." My father, ex-air force (23 years in), who was never impressed with civilian police, said "you need to get off my property before you wind up needing help to walk. You can mail the god damned ticket, or come back at a more presentable hour," then slammed the door in his face. The shithead did come back later that morning (once it was daylight), with another officer "just in case," to hand-deliver the citation, but again my dad just said "he's a minor, I'll sign for it myself, you won't be seeing him today unless you have a warrant you haven't shown me yet." They didn't like it, but they indeed handed him the ticket, which he took (without signing) and then slammed the door in their face again.

We lawyered up, and on the assigned court date, the clerk mysteriously couldn't find any record of a citation having ever been filed (much to the attorney's irritation). There were no charges and no case to be tried/heard. After a bit of digging, we found out the punk just happened to be friends with a local cop, and he called in a favor to "scare me a bit." That might have explained their reluctance to escalate when my dad told them to go fuck themselves, but then again, cops know better than to pull that kind of shit for a simple non-violent non-collision non-criminal misdemeanor moving violation in the first place, and given my father (in his military days) had roughed up some civilian cops for misbehaving in the past he'd have still told them to fuck off even if they were "legally" in the right. But (yet) again, you don't hand-deliver traffic tickets like that, and he knew it.

It's fine if you want to be mad. But don't act helpless. Fight back.

There is no capitalism in this country, just a plutocracy. It's why all of the big American businesses depend on the government-backed monopoly known as copyright. A government will wake up teenagers with guns pointed at them for the crime of harming the larger businesses.
Oh god you're one of those. Just shut up, JFC. Move to a country where you're not cucked by the gun laws and cops know better than to draw guns for civil disputes.
 
Back
Top Bottom