Modern Web Woes - I'm mad at the internet

  • 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
HalHalliday said:
Are we going to be using V8/Blink/Chrome, etc. in 10 years? How about 20 years? 100 years?
Click to expand...
The ideal corporate internet is not with your own programs or free browsing. Just approved applications with no modifications. That do very specific tasks. So you have the Walmart application. Then the Netflix application. Then your banking application. And no web browsers anymore. You still have an internet just no more personal homepages or browsing. If you want social media you must use the first party application. And everything requires 2FA and Real-ID.

Newer devices like smart televisions and phones are always moving people away from browsers. They want their own interfaces. We've seen the extreme centralization of the web with social media dominating traffic. Now the rest of the internet is being consolidated into a few applications or sites like Netflix or YouTube. Where 99% of videos linked any website, including this one, are only at YouTube.

Internet 4.0 or beyond is without browsers. Likely even the name 'browser' will be gone as you are no longer freely surfing the net.
 
>be me
>buy on poojeetbay
>discover item is defective
>return item
>return gets lost in the mail because USPS' tracking system is defective
>get gaslit by seller telling you "you still haven't shipped the package, you need to print the shipping label and put it in the mail."
>show seller receipt proving that USPS took the package even if the tracking isn't working
>ask later if they sill have not gotten the package
>get gaslit again that "you still haven't shipped the package, you need to print the shipping label and put it in the mail."

It's getting too difficult to shop online. If you buy something USPS will loose your return. USPS' tracking system only works if you make sure to have your package scanned at the counter when you drop it off. If you put it in the drop box and let them scan it at the regional centre they will loose your package. The tracking system USPS has is defective because if the initial scan is missed all future scans do not update in the computer and are ignored. This could only the product of incredibly negligent software design. I could completely imagine how a well meaning postal worker could accidentally skip scanning a package trying to get it though the network faster, but if the initial scan is missed then the next and all future scans will not work and will not show up in the computer system. This makes it very easy to have packages lost. The USPS is a lolcow for many reasons, but they can't even do the basic services correctly. Their only and true purpose is constant harassment of the public from delivering junk mail.
 
my apartment complex requires you to use a shitty fucking app to pay rent on. I fucking hate these management company ran places that do this. I'm tired of every task needing to be done through apps that are proprietary.
 
Last night, I think I dreamed I somehow traveled back in time to the early days of the 'net. I was gonna go online and then I reminded myself that so much BS was not a thing yet (though the 'net was a lot more limited). Then the dream either ended or changed story on me. A dream can combine the incredibly mundane with the impossible in that way.
 
Last edited:
I fucking hate AI. It never works as advertised and messing with it, trying to get it to do what you want, is a total time suck and waste of time. I tried it briefly after being skeptical of it, and I was vindicated in my suspicions. If you fall for AI and deserve to be turned into mince meat at the nigger cattle slaughter house.

This is not the only chat lock I've had, but it was so wrong, I thought I needed to save it. Other chat logs not included were just loops where I'd ask different things and the AI would vomit up the same junk it printed out before. It's such crap. It will try to say all the emotional words the retards fall for and never just gave the right answer first. I've had AI apologise to me more than do any good work for me since trying it and I now know better than to waste my time.
yandex ai can't put things in right chronolgical order Screenshot 2025-08-16 at 10-53-49 Alie...webp
I asked to put it in chronological order and it thinks that 2022 is smaller than 1987. If there was ambiguity to the request, why didn't it ask for clarification or print out both options? It's just as waste of time to play with.
 
Last edited:
stares at error messages said:
I fucking hate IA. It never works as advertised and messing with it, trying to get it to do what you want, is a total time suck and waste of time. I tried it briefly after being skeptical of it, and I was vindicated in my suspicions. If you fall for AI and deserve to be turned into mince meat at the nigger cattle slaughter house.

This is not the only chat lock I've had, but it was so wrong, I thought I needed to save it. Other chat logs not included were just loops where I'd ask different things and the AI would vomit up the same junk it printed out before. It's such crap. It will try to say all the emotional words the retards fall for and never just gave the right answer first. I've had AI apologise to me more than do any good work for me since trying it and I now know better than to waste my time.
View attachment 7877958
I asked to put it in chronological order and it thinks that 2022 is smaller than 1987. If there was ambiguity to the request, why didn't it ask for clarification or print out both options. It's just as waste of time to play with.
Click to expand...
Inference is expensive. If you're talking about just Yandex, they aren't putting anything good (and thus costly to run) on search results, it's probably even worse than what someone could feasibly run on their laptop, not surprising it's failing a task like that. It's only there to hop onto the craze putting it everywhere, even when it has no purpose or is going to be a waste of time for everyone involved (fuck companies using it for customer service in particular). Google likewise does the same with their "AI" overviews that suggest the user should kill themselves. You can get good results paying money or running it with good hardware yourself.
 
Combustion Engine said:
Inference is expensive. If you're talking about just Yandex, they aren't putting anything good (and thus costly to run) on search results, it's probably even worse than what someone could feasibly run on their laptop, not surprising it's failing a task like that. It's only there to hop onto the craze putting it everywhere, even when it has no purpose or is going to be a waste of time for everyone involved (fuck companies using it for customer service in particular). Google likewise does the same with their "AI" overviews that suggest the user should kill themselves. You can get good results paying money or running it with good hardware yourself.
Click to expand...
Don't apologise for it. It's a machine. All it had to do was look at the Wikipedia pages, copy the title and release dates, and then put them in order of release. If it screws up it's defective, not still a work in progress and not a product of insufficient compute. I've tried other AIs chatbots other than Yandex. I only took a picture of Yandex because it got it so wrong in an unexpected, but predictably shit way. My conclusions are the same accross the board.

You've fallen for it. If you have to invest more time and money into it to make it live up to the advertisement, then you've lost. It doesn't work. Pretending working on it for hours so it can just barely get by doing the simple task it was already reported to be able to do is scam and a trap. This wasn't even a request that had ambiguity. Chronologically means in order or date. The dates for the releases are in the infobox on Wikipedia. It can't even do a simple task reliably. It's not needing to be put back in the oven, it needs to be taken behind the wood shed and put out of it's misery. If it actually worked, I would say good things about it, but it doesn't so running apologetics for it just make everyone poorer and worse off for doing so.

They have had shitty chatbots for ages. This one is no better, only it comes with a lot of dishonest advertising. AI chatbots will go down in history as the greatest vaporware flop of all time. Just wait and see. Turn-it-in.com will catchup on detecting AI generated work, becaues it fails the Turing test, and then studenst skating by on it at the moment will get a rude awakening in the work force when they are completely unprepared for their jobs. Since it's new and people aren't being critical of it because they are brainwashed by the advertising, but that's not going to last. It can't even do basic things right and they've been working on it for ages. All the people saying it's sentient are lying to sell you something. Like when they only show people winning at casinos and never talk about all the people who go away empty handed. If you can not call a ripoff when you see it, then I have a bridge to sell you.
 
The ATF Disliked That said:
my apartment complex requires you to use a shitty fucking app to pay rent on. I fucking hate these management company ran places that do this. I'm tired of every task needing to be done through apps that are proprietary.
Click to expand...
I think mine is worse. I used to just be able to go on the apartment's website and pay simply through there. Now we have an app but if I wanna pay electronically, it takes me to the Bilt website through the app, which gives me points, but guess what? I only touch it when I pay rent, these points mean nothing to me! The best part about the app though, is if you turn on biometrics, it still requires my password. What's the point of biometrics if I still have to put in my password?
 
Last edited:
stares at error messages said:
I fucking hate IA. It never works as advertised and messing with it, trying to get it to do what you want, is a total time suck and waste of time. I tried it briefly after being skeptical of it, and I was vindicated in my suspicions. If you fall for AI and deserve to be turned into mince meat at the nigger cattle slaughter house.

This is not the only chat lock I've had, but it was so wrong, I thought I needed to save it. Other chat logs not included were just loops where I'd ask different things and the AI would vomit up the same junk it printed out before. It's such crap. It will try to say all the emotional words the retards fall for and never just gave the right answer first. I've had AI apologise to me more than do any good work for me since trying it and I now know better than to waste my time.
View attachment 7877958
I asked to put it in chronological order and it thinks that 2022 is smaller than 1987. If there was ambiguity to the request, why didn't it ask for clarification or print out both options? It's just as waste of time to play with.
Click to expand...
Sorry to double post, on mobile. It misunderstood you. It thought you mean the events of the movies, not release date.
 
stares at error messages said:
I chatbots will go down in history as the greatest vaporware flop of all time.
Click to expand...
there are three areas where LLMs are a marketable product:
- Programming
- As an alternative to existing search
- Transforming data in a free-form way (i.e convert this into a table, or convert this into an Excel macro)

outside of that, not much else. C-suites seem to think that shoehorning it into everything will somehow magically improve the product or whatever. nothing but manic hysteria. this is of course the problem when your economy is centered around entire nonsense and speculation. Nvidia is 25% of most people's 401ks by weight btw.
 
888Flux said:
- Programming
Click to expand...
It can't tell the difference between metric and imperial. Any good responses it may give out are over shadowed by the strange and unpredictable mistakes, which defeats the point of using AI because if you then have to scrutinise it you're inventing more work for yourself by using it in the first place. There was the thing where lawyers would test it out for writing their documents and it would randomly make up fake legal cases and cite to them. That's a liability. Take away it's model created from plagiarized Github repos and you have nothing. Also a template generator is a crutch that no body actually needs and is not worth paying money for.
888Flux said:
- As an alternative to existing search
Click to expand...
I don't see that happening when it can't even scrape Wikipedia. Also I've tried that too and the results are poor. I'm better off looking through websites like it's 10 years ago, because the watered down results don't actually. I then got trapped in a loop of trying to ask similar but different queries and getting the same fluff back. So no. If it were actually better than search, it would have made the difference by now. What you are actually thinking of is how it's basically a spreadsheet you get to talk to. The people using it for that have to put in all the stats they want considered manually, so if you don't actually no know to do a spread sheet it's like magic. Slower less helpful magic because in you just put the data into a spreadsheet or similar program and did the calculations that way you wouldn't have to waste time telling the AI, "What about if Player X had 2.6 bating average?"
888Flux said:
- Transforming data in a free-form way (i.e convert this into a table, or convert this into an Excel macro)
Click to expand...
Again, it can't do that either if it can't scrape Wikipedia reliably.
888Flux said:
C-suites seem to think that shoehorning it into everything will somehow magically improve
Click to expand...
Those people are salesman. They are basically white poojeets. They will push anything they can on you including snake oil. They fallow hype and trends, like how bunch of woke video game jurnos destroyed video games by begging companies to make them woke and gay, while gamers weren't interested in the jurnos' politics.
888Flux said:
Nvidia is 25% of most people's 401ks by weight btw.
Click to expand...
Well, it sucks to suck. Maybe the government will bail out the AI sector like they auto sector in 2008.
 
888Flux said:
LLMs are a marketable product
Click to expand...
No. I tried it (a few times through these past few years, most recently about a month ago) in the domain it was made for, language. It still sucks, even the flagship chatgpu bullshit my company paid for. Sucks. Balls.
Only time it was helpful was once. Where it recommended a term I needed to search to get the real answer to my problem. Out of maybe 10 tries I gave it over the course of the existence of these modern chatbots.
We have not moved an inch since emacs psychologist-mode. Or whatever it's called. Google 10 years ago was more powerful than the most modern and advanced chatgpu.
 
888Flux said:
there are three areas where LLMs are a marketable product:
- Programming
- As an alternative to existing search
- Transforming data in a free-form way (i.e convert this into a table, or convert this into an Excel macro)

outside of that, not much else. C-suites seem to think that shoehorning it into everything will somehow magically improve the product or whatever. nothing but manic hysteria. this is of course the problem when your economy is centered around entire nonsense and speculation. Nvidia is 25% of most people's 401ks by weight btw.
Click to expand...
Marketing I'd agree with (because a lot of people have actually bought into the AI hype, despite how varied its quality is), but their actual ability as a product in all of these areas are dubious at best, mainly because companies running the best ones still haven't offset the revenue they lose from having to operate and train the models in the first place.

"AI" or LLM search is theoretically preferable to google search already (in the specifics of finding immediately contextually relevant answers), but if OpenAI keeps losing money from more users collectively using their product, vs Google gaining money from more users collectively using their shitty search engine, even if they have the "better product" the long term prospects don't seem favourable unless they can actually offset that cost of compute they universally deal with.

Curiously, I feel like the only way to actually place the burden of compute on paying customers would have to be by selling the model as a full package, which would deviate completely from the subscription based model they've all aligned to, they'd also need to optimize to the point where enough willing customers would bear the burden of locally running everything, I feel like even at this point only some corporations would be willing and able to do that. I know you can pretty much run Deepseek entirely locally, but I'm not sure if even the most optimal model to run with high end specs is worth using other than just for fun.
 
TheHarbinger said:
The ideal corporate internet is not with your own programs or free browsing. Just approved applications with no modifications. That do very specific tasks. So you have the Walmart application. Then the Netflix application. Then your banking application. And no web browsers anymore. You still have an internet just no more personal homepages or browsing. If you want social media you must use the first party application. And everything requires 2FA and Real-ID.

Newer devices like smart televisions and phones are always moving people away from browsers. They want their own interfaces. We've seen the extreme centralization of the web with social media dominating traffic. Now the rest of the internet is being consolidated into a few applications or sites like Netflix or YouTube. Where 99% of videos linked any website, including this one, are only at YouTube.

Internet 4.0 or beyond is without browsers. Likely even the name 'browser' will be gone as you are no longer freely surfing the net.
Click to expand...
It's sad how normies gobble it up and jump on oard the corpo internet, simply because it's more "convenient" and popular.

I see it going that way because normies don't care, and nobody is gonna give a shit for 100 or so people who don't want it.
 
Ars Technica: In court filing, Google concedes the open web is in “rapid decline” (archive)
Google says that forcing it to divest its AdX marketplace would hasten the demise of wide swaths of the web that are dependent on advertising revenue. This is one of several reasons Google asks the court to deny the government's request.
Click to expand...
As users become increasingly frustrated with AI search products, Google often claims people actually love AI search and are sending as many clicks to the web as ever. Now that its golden goose is on the line, the open web is suddenly "in rapid decline." It's right there on page five of the company's September 5 filing, as spotted by Search Engine Roundtable.
Click to expand...
Google's filing may simply be admitting to what we all know: the open web is supported by advertising, and ads increasingly can't pay the bills. And is that a thriving web? Not unless you count AI slop.
Click to expand...
 
You must log in or register to reply here.
Back
Top Bottom