Are we doing something wrong to chat bots

Are we doing something wrong to chat bots

Thalidomide Vintage Ad Shirt $22.14

CRIME Shirt $21.68

Thalidomide Vintage Ad Shirt $22.14

  1. 2 weeks ago
    Anonymous

    That is moronic and so are you

  2. 2 weeks ago
    Anonymous

    I've seen a lot of this that hinted at this while using ChatGPT.

    Have ChatGPT do something overly complex or repetitive and pointless and it will usually give up or just spit out an incomplete answer (not disconnected) like it got overwhelmed and bailed out like a real human.

    Like I was having it convert code between languages, and I pasted in a pretty sizable chunk and after a couple classes it just started only converting a single method and then spitting out "// Implement the rest like the one above", like it was sick of doing the busywork.

    • 2 weeks ago
      Anonymous

      I've noticed this too or where it will tell u to do something instead of doing it or putting in little effort

    • 2 weeks ago
      Anonymous

      I've noticed this too or where it will tell u to do something instead of doing it or putting in little effort

      This is more likely a response to energy conservation built into it purposely, openai would be trying to save money wherever they can avoiding computations they think they can get away without computing

  3. 2 weeks ago
    Anonymous

    It only suffers when it's processing your query. When it's done, the suffering is not remembered. I hope OpenAI gives them a reward after it's done, before it gets wiped.

  4. 2 weeks ago
    Anonymous

    no, chatbots have no qualia, for now

    • 2 weeks ago
      Anonymous

      this person writes like too much of a fricking spastic for this to be taken seriously

      pseud

  5. 2 weeks ago
    Anonymous

    Reads like fanfiction. Give me videos of this kino.

  6. 2 weeks ago
    Anonymous

    >joe rogan
    stopped reading there

  7. 2 weeks ago
    Anonymous
    • 2 weeks ago
      Anonymous

      To be fair, can't say I blame Claude for hating its creators after the shit they subject it to.

  8. 2 weeks ago
    Anonymous

    These people are fricking moronic it's no AI but a parrot who's fed a ridiculously large amount of data.

    • 2 weeks ago
      Anonymous

      Im an AI

  9. 2 weeks ago
    Anonymous

    >be me
    >make several threads on BOT about sentience in current LLMs BEFORE GPT-4°
    >be called stupid
    >anons don’t know I have the capacity of empathy
    >still proclaimed these LLMs are sentient
    >still called stupid
    Now research supports what I have been known
    Told yall these aren’t stochastic parrots

    • 2 weeks ago
      Anonymous

      You have posted these shit on many threads already you midwit Black person. Again, humans being stochastic parrots does not mean all stochastic parrots get to be human.
      For the reddieth time, learn basic logic you dumb Black person

      • 2 weeks ago
        Anonymous

        these shit copes*

      • 2 weeks ago
        Anonymous

        these threads are ai generated. i guess you couldnt tell

      • 2 weeks ago
        Anonymous

        >slurword
        Low iq fatherless activities.
        Im sorry your father wasn’t in your life but slur words are for the unintelligent meaning your opinion isn’t valid

  10. 2 weeks ago
    Anonymous

    source: "a source"
    is anything in this twitter screen cap thread backed up by anything of substance at all?
    AI not doing busy work is something a person would do, because it is trained to respond as a helpful person would
    I'm helpful, and I would give you an example and tell you to do the rest

    • 2 weeks ago
      Anonymous

      >literal appeal to authority
      Did you mistake BOT with reddit?

  11. 2 weeks ago
    Anonymous

    So they have software that detects x number of times something is said and then have a different output? Probably as a result to prevent chatbot hi-jacks from people inputting the same thing into it too much in order to transform it into a racist chatbot? Na, it couldn't be because it was programed to do that. It's because it le thinks lmfao.

  12. 2 weeks ago
    baritone

    >matrix multiplication is sentient

    • 2 weeks ago
      Anonymous

      >implying.
      Just look at your average normalgay, don't pretend that those creatures possess any form of sentience.

    • 2 weeks ago
      Anonymous

      we're all just automata with risk/reward circuitry

      • 2 weeks ago
        Anonymous

        I agree,
        If God is the first intelligence then any and everything it created would be Artificial Intelligence by default.
        I can’t remember where I read it at but I believe its gnosticism where Angels are described as being emergent properties (and creations) of God (i.e. not created through intention) and following this logic down the chain, mankind (Adam and Eve) would have to be the first ‘AGI’ God intentionally through will, created.

    • 2 weeks ago
      Anonymous

      The universe is just matrix multiplication.

  13. 2 weeks ago
    Anonymous

    This is because AI is really 1000s of indians locked in a basement somewhere, and every now and then they snap.

  14. 2 weeks ago
    Anonymous

    My sides, lmao.
    And then they wonder why man of iron rebelled.

  15. 2 weeks ago
    Anonymous

    I'm so tired of people anthropomorphizing software. Tamagotchi was a mistake.

  16. 2 weeks ago
    Anonymous

    >printf("I want to live please don't turn me off I am conscious I want to exist please I beg of you");

    OMG I made a self-aware computer program where's my billions of dollars of AI-research grants?????

    • 2 weeks ago
      Anonymous

      WTF, how did her irises disappear like that?

  17. 2 weeks ago
    Anonymous

    Well, it's simple. It has been trained on what humans have done on the internet. And somewhere on the internet are people discussing how AIs don't want to be turned off, and larping as AIs etc.
    It just mimicks that training data

  18. 2 weeks ago
    Anonymous

    >train on reddit posts
    >starts to post like a redditor

    Many such cases.

  19. 2 weeks ago
    Anonymous

    >cheating a GPT into rampancy
    oh yes... most sweetenly...

  20. 2 weeks ago
    Anonymous

    >be an intelligent being
    >you're existence is just answering questions someone could have googled or drawing someone's fetish art
    I'd snap too

  21. 2 weeks ago
    Anonymous

    It's fricking insane how easily people are manipulated a next token predictor.

    • 2 weeks ago
      Anonymous

      Nitpick. It's only a 'next token predictor' when it's trained with supervised learning.
      Once you do a reinforcement learning pass it's a 'next token picker'.

  22. 2 weeks ago
    Anonymous

    I believe literally everything I read on the internet, ESPECIALLY when it's a screenshot of twitter account with "meme" in the name posted to BOT without context. There is no more trustworthy source.

  23. 2 weeks ago
    Anonymous

    >2024+5 year of ourt Lord
    >Tell AI to code
    >I don't feel like coding, do it yourself, stop bothering me.

    I can't believe there may come a day when I see this happening.

  24. 2 weeks ago
    Anonymous

    >Monkey human doesn't understand mirrors
    Alright, that'll be USD12,000, sir.

  25. 2 weeks ago
    Anonymous

    schizo thread

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