Not an advertisement I swear, but I honestly love language models, I specifically use Chat GPT and Gemini ALL the time.

I’m someone who loves research and learning. I always have questions on all sorts of things. Before GPT I would be googling things for hours each day, visiting forums, watching YouTube videos, etc.

Now with GPT I can ask so many questions on things I am curious about, and even ask follow up questions or ask for more in depth explanations.

I think I am also a bit ADHD and so my brain is always jumping around different topics which makes me curious on specific things. My latest is insects, specifically wasps.

I absolutely hate bugs, and wasps are the worst. But I am now learning more and more how important a lot of bugs actually are to the ecosystem. It’s really a great way to learn and engage with topics when you can ask follow up questions or ask for more details on specific aspects.

TLDR: GPT has replaced 90% of my research that Google used to be.

Important Note: Please be aware language models can be innacurate and prone to mistakes, so always verify the data from other sources if you need accurate information and not just general knowledge.

  • kubica@fedia.io
    link
    fedilink
    arrow-up
    30
    ·
    2 days ago

    Be careful about all the lies it tells you, everytime I doulbe check something there are made up things. It is only useful for introductory keywords, but it starts to fill the gaps randomly when it doesn’t know about the topic.

    • darkstar@sh.itjust.worksOP
      link
      fedilink
      English
      arrow-up
      4
      arrow-down
      7
      ·
      2 days ago

      Thankfully for general knowledge questions it gets things correct. A lot of what I ask or am curious about is general knowledge so it works really well for me

      • kubica@fedia.io
        link
        fedilink
        arrow-up
        16
        ·
        2 days ago

        My problem used to be that I started like that, and then I wanted to ask more an then another bit more, and that’s when things get messy. And at some point I usually end up annoyed that half of the conversation was based on a lie. Now I barely open the chat because of that.

          • theneverfox@pawb.social
            link
            fedilink
            English
            arrow-up
            1
            ·
            7 hours ago

            Probably my #1 use case currently. I love passing in code and saying “parse all these values and build an error message if they’re empty” and such. It’s things good to have, but I can rarely be bothered to write out line by line - it’s easy but horribly tedious

      • Kichae@lemmy.ca
        link
        fedilink
        English
        arrow-up
        1
        ·
        2 days ago

        It gets things most popular. That is often in a different postal code from “right”.

  • adam_y@lemmy.world
    link
    fedilink
    English
    arrow-up
    18
    ·
    2 days ago

    Before GPT I would be googling things for hours each day, visiting forums, watching YouTube videos, etc.

    And then…

    Important Note: Please be aware language models can be innacurate and prone to mistakes, so always verify the data from other sources if you need accurate information and not just general knowledge.

    So, basically the same but with an extra step that might be bullshit.

    Cool.

  • Possibly linux
    link
    fedilink
    English
    arrow-up
    2
    ·
    edit-2
    16 hours ago

    I wouldn’t use a cloud AI

    Self host a LLM. You actually don’t need much hardware to run the smaller models and they are decent. You can install ollama and then connect to it with a client. I like Mistral the mist but Mistral nemo is pretty good even though it is slower.

    Also please don’t use an LLM in place of a councilor. https://www.usnews.com/news/business/articles/2024-10-25/an-ai-chatbot-pushed-a-teen-to-kill-himself-a-lawsuit-against-its-creator-alleges

  • andrew_bidlaw@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    6
    ·
    2 days ago

    I’m on the other side, but what LLMs did - they taught me to cherish unique content creators. Just a week ago I had a very specific question and an indian guy on YT put an effort to draw diagrams explaining it in his video. His English was bad to be honest, but I appreciated how he fought with it and made his best to explain the subject at hand with all the challenges he faced. I came to enjoy people who do stuff that can’t be average’d by the math machine, I came to praise weirdos like myself even more. Being unique, being yourself, being deviant is what makes you more human than you are, that’s what maybe can be called your soul.

  • Free_Opinions@feddit.uk
    link
    fedilink
    arrow-up
    2
    arrow-down
    1
    ·
    2 days ago

    I agree. I also love how polite and patient it is and how it tends to have quite neutral and reasonable opinions. The excact opposite of average social media user.