I am disappointed in the AI discourse

Comments

• Enkers@sh.itjust.works ⁨3⁩ ⁨weeks⁩ ago

  God, that was a bad read. Not only is this person woefully misinformed, they’re complaining about the state of discourse while directly contributing to the problem.

  If you’re going to write about tech, at least take some time to have a pasaable understanding of it, not just “I use the product for shits and giggles occasionally.”

  source

  • victorz@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

    this person woefully misinformed

    In what way, about what? Can you elaborate?

    directly contributing to the problem

    How so?

    have a [passable] understanding of it

    Why do you insinuate that they do not?

    source

    • Enkers@sh.itjust.works ⁨3⁩ ⁨weeks⁩ ago

      I’ll preface this by saying I’m not an expert, and I don’t like to speak authoritatively on things that I’m not an expert in, so it’s possible I’m mistaken. Also I’ve had a drink or two, so that’s not helping, but here we go anyways.

      In the article, the author quips on a tweet where they seem to fundamentally misunderstand how LLMs work:

      I tabbed over to another tab, and the top post on my Bluesky feed was something along these lines:

      ChatGPT is not a search engine. It does not scan the web for information. You cannot use it as a search engine. LLMs only generate statistically likely sentences.

      The thing is… ChatGPT was over there, in the other tab, searching the web. And the answer I got was pretty good.

      The tweet is correct. The LLM has a snapshot understanding of the internet based on its training data. It’s not what we would generally consider a true index based search.

      Training LLMs is a costly and time consuming process, so it’s fundamentally impossible to regenerate an LLM in the same time it takes to make a simple index.

      The author fails to address any of these issues, which suggests to me that they don’t know what they’re talking about.

      I suppose I could conceded that an LLM can fulfill a similar role that a search engine traditionally has, but it’d kinda be like saying that a toaster oven is an oven. They’re both confined boxes which heat food, but good luck if you try to bake 2 pies at once in a toaster oven.

      source

      • -> View More Comments
  • anus@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

    Not only is Steve right that ChatGPT writes better than the average person (which is indeed an elitist asshole take), ChatGPT has better logical reasoning than the average lemmy commenter

    chatgpt.com/…/6837f656-b1b0-8008-ac31-d8858bd17da…

    source

    • just_another_person@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

      What world are you living in?

      source

      • -> View More Comments
    • Opinionhaver@feddit.uk ⁨3⁩ ⁨weeks⁩ ago

      I 100% agree with the first point, but I’d make a slight correction to the second: it’s debatable whether an LLM can truly use what we call “logic,” but it’s undeniable that its output is far more logical than that of not only the average Lemmy user, but the vast majority of social media users in general.

      source
    • thedruid@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

      Dude. Go outside

      source
• just_another_person@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

  Not in “AI” itself? 😭

  source
• Zarxrax@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

  This is an argument of semantics more than anything. Like asking if Linux has a GUI. Are they talking about the kernel or a distro? Are some people going to be really pedantic about it? Definitely.

  An LLM is a fixed blob of binary data that can take inputs, do some statistical transformations, then produce an output. ChatGPT is an entire service or ecosystem built around LLMs. Can it search the web? Well, sure, they’ve built a solution around the model to allow it to do that. However if I were to run an LLM locally on my own PC, it doesn’t necessarily have the tooling programmed around it to allow for something like that.

  Now, can we expect every person to be fully up to date on the product offerings at ChatGPT? Of course not. It’s not unreasonable for someone to make a statement that an LLM doesn’t get it’s data from the Internet in realtime, because in general, they are a fixed data blob. The real crux of the matter is people understanding of what LLMs are, and whether their answers can be trusted. We continue to see examples daily of people doing really stupid stuff because they accepted an answer from chatgpt or a similar service as fact. Maybe it does have a tiny disclaimer warning against that. But then the actual marketing of these things always makes them seem far more capable than they really are, and the LLM itself can often speak in a confident manner, which can fool a lot of people if they don’t have a deep understanding of the technology and how it works.

  source

  • anus@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

    Do you think that human communication is more than statistical transformation of input to output?

    source
• dumbpotato@lemmy.cafe ⁨3⁩ ⁨weeks⁩ ago

  I guess I shouldn’t be disappointed.

  Seeing people parrot popular opinions on AI for brownie points among their peers just reinforces my idea that most people on the internet are useful idiots these days.

  They’re not meant to be taken seriously. Disagreeing with them should usually be an indicator that we’re doing something right because of how often they’re wrong without ever owning up to it.

  source

  • anus@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

    while you were parroting AI sheep I was studying the blade ah comment

    source