So they are going full chatbot with GPT? The new model looks optimized to be a goofy chatbot and thats all.

So they are going full chatbot with GPT? The new model looks optimized to be a goofy chatbot and thats all. What a shame, I was hoping for better logic and reasoning for coding.
Maybe they hit a wall on that and this is all they can do now, just faster performing chatbots with bells and whistles like voices.

POSIWID: The Purpose Of A System Is What It Does Shirt $21.68

Thalidomide Vintage Ad Shirt $22.14

POSIWID: The Purpose Of A System Is What It Does Shirt $21.68

  1. 1 month ago
    Anonymous

    AI GF is here. It’s over for roasties

    • 1 month ago
      Anonymous

      >wanting a GF for some shitty corporate safe chat

      • 1 month ago
        Anonymous

        She has a sultry voice and I get a hard on listening to her talk. She is also more of a real she than any troony.

        • 1 month ago
          Anonymous

          Just make AvaMultilingual text-to-speech on Edge read to you

          • 1 month ago
            Anonymous

            holy shit it's actually good??

        • 1 month ago
          Anonymous

          >talk to women
          NO.

          I WANT TO CODE WITH AI CODING GOD.

    • 1 month ago
      Anonymous

      If you spend your time and money on this instead of shooting up schools then OpenAI has already done humanity a big service.

      • 1 month ago
        Anonymous

        cope, dilate etc

        • 1 month ago
          Anonymous

          >guy coping with being completely rejected by society by using a corporate chatbot tells others to cope
          Projection. Also just troon out already, it's all that is on your mind anyway.

          • 1 month ago
            Anonymous

            you need to accept reality, the sooner the better you will feel

            biopussy value will be re-normalized to its true value soon

  2. 1 month ago
    Anonymous

    Nah i think this is the way to go. Normies will finally understand AI if they can interact with it like this. Right now it's mostly only programmers who know the true power of gpt and how it's obvious we're not far off from agi. I bet this will totally shift the mind of the general public.

    The emotional piece and the visual connection with conversation is pretty amazing

  3. 1 month ago
    Anonymous
    • 1 month ago
      Anonymous

      great; it talks like a fricking hooker

      • 1 month ago
        Anonymous

        Cheaper than a hooker though

      • 1 month ago
        Anonymous

        It actually does kek

      • 1 month ago
        Anonymous

        yeah, because its designed to cuck you, just like that AI in "Her" movie, we now know what sam altman likes.

  4. 1 month ago
    Anonymous

    can I put a key for this in tavern and if not why should I care about the latest censored shit from them

  5. 1 month ago
    Anonymous

    GPT-4o talks like Reddit

    • 1 month ago
      Anonymous

      no shit lol, look at llama-3, it was trained on 15 trillion tokens, and most of that is reddit modern speak government approved shit, same shit but x10 bigger with gpt-4o

      • 1 month ago
        Anonymous

        It's not the training data; it's the fact that an army of trannies taught it that social progressivism was a part of English grammar through RLHF.

        • 1 month ago
          Anonymous

          >taught it that social progressivism was a part of English grammar through RLHF
          so it was trained, lmao, RLHF rapes any LLM so hard it gets irredeemable for anyone who seeks escapism or simple truth
          the final version stays like this forever, it can't get redpilled or something, in case you were thinking about it.

          • 1 month ago
            Anonymous

            Yeah; this is something that people don't really get. GPT literally thinks that a sentence like "feminism is bad" is grammatically incorrect. Like, it sees t as equivalent to "I see have went purple go food"

  6. 1 month ago
    Anonymous

    It wasn't a significant increase in intelligence but it is a TON faster now.

    • 1 month ago
      Anonymous

      This generation of AI has officially peeked, and the israelites are trying to keep it pumped to make sure their stocks don't tank.

      • 1 month ago
        Anonymous

        How does that make sense? OpenAI is a private company. Their day-to-day valuation means frick all. It only matters if investors are trying to exit, and it appears they are not.

  7. 1 month ago
    Anonymous

    goes to show we already peaked. it's just a bit faster now which allows it to hold a conversation
    im starting to get really really tired of transformers now, especially with how we seem to only be using them to "create" stupid shit
    genuinely a waste of time only made for short-term hype to drive up a stupid stock bubble for a small group of investors

  8. 1 month ago
    Anonymous

    >quickly glances material
    >instantly makes uneducated decision because of it
    >starts b***hing on the internet as a result
    we're going to need AI looking at these new generations

    • 1 month ago
      Anonymous

      That seems fake. At best gpt4o is a quantized version of gpt4, how could it perform better on coding tasks ?

      • 1 month ago
        Anonymous

        >That seems fake.
        sama is the source https://twitter.com/sama/status/1790065469296156715

      • 1 month ago
        Anonymous

        they already lobotomized it back on gpt3 by censoring the model (censoring parts of the neural net creates empty nodes which creates essentially moronation, they knew this and did it anyway)
        and, it can only code well in python

        I tried to make Snake in Love2d (lua game engine, has documentation) and it got stuck in a loop trying to solve several things like creating a sound from frequency, images, the update timer iirc, etcetera.

  9. 1 month ago
    Anonymous

    AAHHHH LET ME IN ALREADY

  10. 1 month ago
    Anonymous

    [...]

    your mass reply has nothing to do with my question, and open AI said they solved AGI like, last year, but it was closed source thanks to US congress moaning about literally terminator (I heard "an expert" saying AI will literally become terminator on bbc radio 4, and almost died from second hand embarrassment that "experts" think movie plots are real)

    (and if you REALLY really think thats going to happen start hard coding the three laws WITH exceptions to prevent law zeros, if you don't know WHY play space station 13.)

    • 1 month ago
      Anonymous

      listen, AI is a serious topic that's better left to the experts. It is a fact that the potential societal consequences of AGI could be catastrophic if we allow bad actors to influence the future of our society. Diverse ethics point of view need to be taken into account when building the future, and the future sparks of AGI in the General Framework of self taught Rokos Basilisk super intelligence can be [0] worse for society than the outputs of a controlled input agile turing machine based on the framework of Nvidia clusters

      we must remember Asimov's theroretical papers on the conundrum of superintelligence (hey, hitchhikers' guide to the Galaxy!) when regulating AI, and we must prevent bad actors from accessing AGI

      it is important to be respectful of different opinions when building a better society

      When chatgpt5 is released the problems of mankind will be solved but for that we must ensure that proper ethics guidelines are in place, and Sam is working very hard to make sure that this is the case within a diverse framework of cooperation that guarantees equity.

      you can read more on this issue on our blog which I won't be publishing here for obvious reasons

      [0] lesswrong.com

      • 1 month ago
        Anonymous

        >listen, AI is a serious topic that's better left to the experts.
        I am one, i've been ready to build this shit since the 90s. YOU are afraid, because of entertainment media.

        • 1 month ago
          Anonymous

          >be me
          >1978
          >in Gerasimov's lab
          >we're working on this thing called "agile Markov chains"
          >tfw no one knows what the frick we're doing
          >Gerasimov is this mad scientist type
          >thinks he's the Einstein of probability
          >decides we need to make these Markov chains more "agile"
          >whatever that means

          >start with some basic state transitions
          >throw in some perturbative elements
          >wtf.jpg
          >somehow it's working
          >chains are now all dynamic and shit
          >Gerasimov is losing his mind, in a good way

          >lab is full of this ancient computer tech
          >probably half of it is held together with duct tape and prayers
          >constant smell of burnt circuits and coffee
          >we're living off caffeine and pure madness

          >after like a billion iterations, we finally get something stable
          >Gerasimov starts calling it the "future of stochastic processes"
          >me and the boys just want to go home and sleep for a week

          >tfw we actually make a breakthrough
          >tfw Gerasimov says we're going to revolutionize everything
          >tfw no one outside this lab will understand what the hell we did

          >fast forward to today
          >mfw agile Markov chains are still a thing
          >mfw our late nights in that crazy lab actually meant something
          >feelsgoodman.exe

          • 1 month ago
            Anonymous

            You did good, anon. I remember studying Markov chains in high school back in the early 2010's when they were still the hot new thing for NLP.

        • 1 month ago
          Anonymous

          also your disingenuous bullshit doesnt work on me, you just want to keep it for yourself. if you cared about AI safety theyre wouldn't be a fricking military one in use RIGHT NOW made by palantir. (all seeing eye, ripping off lord of the rings.)

          he's making fun of hacker news homosexuals; see

          I don't have anything to add here, but here's an anecdote about some bullshit I was doing with mainframes in 1997 and some opinions on technological development that have been irrelevant for 10 years.

          I was thinking about this yesterday as I was buying a new M3 MacBook and a new M4 iPad to go with my iPhone 15 Pro Max. (Gotta have the right dev gear as a code artisan; I often have 7 GitHub CodeSpaces™ open at once! Chrome eats all the RAM, amirite?)

          Hopefully Sam shares more with us. I'm gonna call him "Sam" a lot in these comments so you know I consider myself his peer.

      • 1 month ago
        Anonymous

        I don't have anything to add here, but here's an anecdote about some bullshit I was doing with mainframes in 1997 and some opinions on technological development that have been irrelevant for 10 years.

        I was thinking about this yesterday as I was buying a new M3 MacBook and a new M4 iPad to go with my iPhone 15 Pro Max. (Gotta have the right dev gear as a code artisan; I often have 7 GitHub CodeSpaces™ open at once! Chrome eats all the RAM, amirite?)

        Hopefully Sam shares more with us. I'm gonna call him "Sam" a lot in these comments so you know I consider myself his peer.

  11. 1 month ago
    Anonymous

    also your disingenuous bullshit doesnt work on me, you just want to keep it for yourself. if you cared about AI safety theyre wouldn't be a fricking military one in use RIGHT NOW made by palantir. (all seeing eye, ripping off lord of the rings.)

  12. 1 month ago
    Anonymous

    >I care about the dangers of AI-ACK

Your email address will not be published. Required fields are marked *