'AI is Killing the Old Web' - Slashdot

Posted by msmash

• ›

  ... it's people who can't resist the siren song of bullshit generation that are killing the Web.

  • 1 hidden comment

  • • Re: It's not AI

      Nothing is killing the old web faster than bullshit AI is doing this to sensationalist headlines. Enshitification of TikTok? Really? Since when did rants make stories? Oh right. Since the mention the nest click-bait bingo tag. Twitter is what you make of it, now that the moderators are checked, and bots can be useful if you have them provide something in context that is worth reading to you. Just depends on what you are looking to get out of it.
      • 1 hidden comment

  • Re:

    You cannot stop web 3!!
    All posts will be NFT's owned by the user that created them not by the faceless corporations! And everyone will be able to trace your posts to you as well! What could go wrong! It's totally not a scam!

  • Re: It's not AI

    Halfway true-

    Megacorporations have discovered that there's more money to make from AI generated content than user generated content.

    Why pay actors when you can have an AI? All people need to be is consumers,

    Just go back and re-watch Max Headroom.

• Pockets of the internet not filled with AI generate garbage will survive and thrive. The hype will die down and we'll all move on to the next fad.

  • 1 hidden comment

  • Re:

    Trying to "exclude AI" is going to work as well as trying to "exclude technology". It's going to be in most of the tools you use, and in many cases, you're not even going to realize you're using it. A tool as powerful as AI is not just going to "die down", as much as you hope to wish that into happening. A couple days ago in a I wrote a tool, running on my own computer with a consumer-grade GPU, to summarize a page of text per second. Which (now that I have over 100k samples) I plan to use to train a mode

    • 2 hidden comments

    • Re:

      Yes. Exactly this. Just like the lawyers let AI reference case law.

      It's not that I'm a Luddite but what good is it when you can't really be sure if it's accurate. Just like the self driving car and the automated attendant. It a long way from what is advertised.

    • Re:

      We've had AI since the 1950's. Transformers are neat, but so was Eliza... and expert systems, RNNs, GANs, CNNs, Watson, AlphaGo, DALL-E,... you get the idea. This time is no different than all of the others.

      We've already gone from "ChatGPT is going to end college writing assignments" to "well, maybe students can use it to give them a place to start". Remember when GTP-2 was "too dangerous" to be released to the public? It's all very silly.

      Wishes and hopes are for the true believers. Haven't you noti

• "Twitter is being abandoned"

  No-one goes to Twitter anymore, it's too crowded...

  I've seen zero sign people are abandoning Twitter, my own observations are that people use it more than ever, especially around news events.

  There were a handful of people that left for Mastodon, but most of those came back once they actually got a taste of what it was like to use Mastodon.

  I totally disagree with the broader assertion that AI is killing "The Old Web". What it's killing is clickbait news sites. Slashdot for instance is very "old web", what has changed here? Nothing.

  • 1 hidden comment

  • • Re:

      Why do you care what you are modded to here, people can still read your comments, I read at -1 its fine, sure you get a few trolls posting but I'm an adult I can handle it. If you want your post curated for you sure read at what ever level you choose.

      • 1 hidden comment

    • Re: Abandoned?

      Please explain this concept of "factually based opinions". Cos if it's a fact, it's not an opinion; if it's an opinion it's not necessarily a fact. You can't choose your own facts but you can change your opinion. Well, at least people used to. Stubbornness has become a virtue.
      • 1 hidden comment

      • • Re:

          What's your experiment ?

  • Re:

    I'm really enjoying Twitter since the censorship stopped. It seems the people who bad mouth Twitter are the kinds of people who would bad mouth any non-censored platform.

• As someone who doesn't go to the big social media sites, I have to wonder, how bad is it really? I've been expecting a botpocalypse within the next few years, but is it really here?

  I think probably the saddest outcome will be people who still hang around on LinkedIn or Bookface, knowing they're talking to bots. Somehow I don't expect the sites to shrivel and die entirely, but hopefully enough people will be motivated to pull the jack out of their brainstem that I'll no longer be expected to have an account.

  • 1 hidden comment

  • Re:

    I still use both faceboot and twitler and frankly they have really not changed that much from my perspective. With facebook though I spend almost all my time in private groups, and I am not an independent journalist or a trans person so the changes on twitter don't march up into my face and harass me.

• /. "editor" cites The Verge for his one article a week. Meanwhile, The Verge is the canonical website for killing journalism. Junk writers, zero expertise, yet apparent experts compared to/. editors.

  AI isn't enabling anything that wasn't already happening. SuperKendall has been killing the internet for a decade.

• Out of all the website examples, the submitter forgot to include Slashdot in the list. It's the standard once per post, right?

  The listed examples seem to be modern social media or major websites where the top-executives are letting things rot. Twitter got taken over and became a cesspool, and so on. Almost all of them have to do with breaking down, not from AI but because of mismanagement. As of now, a portion of those sites are being replaced by different options - perhaps some new ones.

  The only bit of concern is AI's self-feeding a misinformation loop, and that can be stopped simply by not using AIs as a reliable source - something that should have been learned when the lawyer cited ChatGPT's non-existent cases. It also impacts Ai-generated pictures, where the feedback loop greatly reduces the quality.

• The AS (artificual stupidity) generated seas of junk misinformation will only make actual precise and correct information more valuable, at least to those who actually need it. The rest were misinformed anyway, and will continue to be so.

• But come on, TikTok was always fake bullshit. It was a CCP social credit system hack full of deepfake pseudo-human faces doing nothing. And now it's even less than that.

  • Re: You had me at "enshitification."

    Enshitification is a defined process which seems to be predictable and also explains the curious behaviours of the biggest internet sites, tiktok inclided

    • Re:

      I'm not familiar with any technical sense of the term, but I agree that it's completely predictable. Fakeness is predictable, and increases like entropy.

      • Re: You had me at "enshitification."

        Not sure if slashdot censoring my comment or what but the source is here absolutely well worth the read Its not about fake ness https://pluralistic.net/2023/0... [pluralistic.net]

        • Re:

          I see the term "Potemkin AI" in the link, and that sounds promising. Hits a lot of points I've already thought myself. Can you give a synopsis first?

          • Re:

            Can you just take your mind off of speaking before thinking and just read the article?

• 'AI is Killing the Old Web'

  I'd call the Old Web the Geocities times, and it's dead in that form for long already. What AI might be killing is the "web 2.0".

  Twitter is being abandoned to bots and blue ticks. There's the junkification of Amazon and the enshittification of TikTok.

  I don't see what AI has to do with that. Maybe it intensified a bit some of it, but the main features (fake users on twitter, junk and paid review on Amazon, and "shit" is almost the very purpose of TikTok -- brainless entertainment for teens) all predated the AI explosion.

  "AI is tearing Wikipedia apart."

  No it's not. Here is the original article that used this expression https://www.vice.com/en/articl... [vice.com] and it becomes clear after reading they just made it a clickbait title. Wikpedia is specifically designed to handle this case. Everyone can write and there is no verification of credentials; it also works for AI. The criterion is humans will read it, and check if there are references and if those references say what the text claims. It does not matter if it was an adult, a child, a trucker, a PhD, or an AI that wrote the text.

  The word "tear apart" is an exaggeration. reality is there are people at Wikipedia who want to use AI tools for certain things (summarizing), others who think we should not. This is not "tearing apart", there are such disagreement on a constant basis at Wikipedia, and they get resolved by consensus-designed policies and community votes.

  To the contrary, the ease to create AI fakes means that now more than ever one needs trustable sources of information. For some, it will be a particular reputed news outlet; for others, it will be a Wikipedia page. More than ever, AI reinforces the role of Wikipedia in providing trustable information.

  • Re:

    Wikipedia does not provide trust-able information; It provides only the information, true or not, permitted by its huge team of politically biased article squatters. At best for anything even slightly political it provides an opinion and a list of sources which tend to support that opinion.

• Happen every few years. I was part of the original enshittification, one of the hordes of people that flooded the internet in 1991, from aol. It was so fucking bad it gas named the "eternal September."
  Well, this is just v3.0 or so, lol.

  • Re:

    Indeed. Eventually it will adjust, but some web things we know and love will get flattened in the process.

    If your site has too much crap, most people will eventually stop visiting it. It may take a year or so until investors/owners get a clue, but except for some niches, you can't have too much crap and survive. The investors/owners will adjust.

    Some sites already tried paying cheap desperate 3rd world workers to generate crap content using semi-AI for various reasons a decade or so ago. It worked to some ex

• If AI makes the internet totally shit, maybe people will stop using it as much and start turning to interacting with real people. And hopefully newspapers, actually newspapers and broadcast news make a comeback where readers/watchers have more of an impact based on how much they subscribe (keeping them honest... if they aren't good no one will buy). Etc etc etc. Make the internet shit to save the world.

  • Re:

    At least you're doing your part to make it happen!

• Years ago, Usenet era, there was a radio program where some then-notable figures were interviewed. Someone offered (paraphrase) that "Usenet functions as a dissent amplifier." Which site will be first to collapse by one LLM trying to out-flame a second LLM, and how likely will it be that neither was designed specifically to wage flamewar?

• Thus by our own actions we create the tools to limit human intellectual capacity to a peak that can't be surpassed. The rising of a process of recycling what is already known has begun and will eventually answer all questions... These tools and process are named artificial intelligence (AI).

  Seriously, I just hope decent news agencies stick to reporting what's actually happening in the world, universities stick to experimenting as part of the course and the places on the web that truly are references continue to be updated by those with actual knowledge. Otherwise the first paragraph is true. AI ends human intellectual progress.

• The Internet was dead years ago, and SEO killed it. They're just using AI as a scapegoat. I've been looking for an out a while, and while I'm ashamed to admit how long it took me to get turned on to them, y'all should check out "awesome lists". It's basically curated content, and some of the interfaces to it have a kind of gopher-ish feel. Nobody owns that phrase, and there are all kinds of lists but they seem to be heavily tech focused for now (that's not a bad sign, it's still got that early feel). What remains to be seen is if it can resist populariztion; but for now it's obscure enough to avoid entanglements with The Empire. Pray they don't alter the bargain any further.

• I don't have much concern about the purported demise of some of the major social media sites but the polluting of them with AI is troubling. This seems very early for that to be happening. I'm concerned that soon it will be impossible to distinguish between humans and AI bots anywhere online, slashdot included of course.

  What's even more troubling is that these LLM's will and are being fed input that's scraped from all the sites where people are posting their human content, whatever it may be. Anything the s

  • Re:

    That sounds like you're describing humans in general. Everything we do is based on plagiarism, from us learning our language to music and arts. You paint a building from an existing building, you make new music based on music you've heard before, you give a stone some eyes - because you've seen a stone and eyes before and now you have a cartoon character. A big youtuber eagerily watches smaller unknown youtubers to steal ideas for the next production, so does the big companies parasiting of others innovati

    • Re:

      There's a specific definition for "plagiarism", and it is not the same as learning from the corpus of human knowledge and building new creations from it.

      https://en.wikipedia.org/wiki/... [wikipedia.org]
      "Plagiarism is the fraudulent representation of another person's language, thoughts, ideas, or expressions as one's own original work."

      • Re:

        It's a matter of scale. âoeTo steal ideas from one person is plagiarism; to steal from many is research.â, attributed to Wilson Mizner
• Not AI. Google had been getting less useful for years, they don't have a carrot in front of them to make their service useful, they can tax the internet with no effort. They get money for being the defacto Internet gatekeeper by forcing companies to have the first link or become irrelevant. Both users and companies suffer

• Google, Twitter, Amazon and such, that is not the old web, but is web 2.0. Old web web was (still is?) personal websites.