Skip to content

Comment on What AI chatbot do you use and why? And some other AI questions

You are viewing one comment and its replies. View all comments

Comments

  • I do not use any AI chatbots. I have ethical, sociological, religious, and environmental reasons for never ever using them.

    If I’m isolating myself by opting out, so be it. I do not want to cook the planet for fucking markov chains.

    5
    • Furthermore, I actively look down on anyone who uses them. I regard them with humanity, as I watch their intelligence and independence and capacity for critical thought dwindle until they’re nothing more than a shell who instinctively reaches for some slop machine when presented with anything of any difficulty.

      My brother was once a masterful artist. He is now nothing more than a meatbag with ChatGPT access, and he is incapable of creating what he made just five years ago.

      3
      • My condolences about your brother. I am noticing a few of my loved ones are in the early stages of losing their capacity for reasoning and learning to these chatbots. It is a horrifying and heartbreaking thing to witness.

        2
    • Meanwhile, I am driven first and foremost by my desire for correct information, but have reached the same conclusion.

      1
    • a markov chain is a very specific and different thing

      0
      • LLM inference fits the definition a Markov chain, where the state is the input string to the LLM and the LLM’s output is taken to be the probability distribution for the next state transition. The state space is the finite set of token strings fitting in the LLM’s context window. The process of selecting the next token to output thus follows the Markov property. A whole readable string of text is generated by iterating this Markov process until the current state ends in an end-of-string token (or some other method).

        As an aside, this is one major reason why people say thay LLMs can’t learn. They are memoryless. Their output is wholly derived from the input state in a deterministic manner (barring inconsistencies in hardware or execution environment) that does not change after training (which is a separate process from inference).

        0
        • ok but basically anything is markovian if you include enough in the state, which doesn’t feel like it fits with what op was trying to imply

          also I’d argue sampling makes it non deterministic

          0

Sign in to comment. Sign in