User Controls

AI general Thread

  1. Originally posted by vindicktive vinny one can just stop taking everything they see on the telescreen seriously.

    Do you have a television vinny?
  2. Originally posted by Jiggaboo_Johnson Do you have a television vinny?

    yes
  3. Originally posted by vindicktive vinny yes

    It's a 2000 era plasma screen isn't it...
  4. Originally posted by Jiggaboo_Johnson It's a 2000 era plasma screen isn't it…

    jokes on you.

    its an early 2010s model.
  5. https://www.livemint.com/technology/tech-news/gpt4-devises-plan-to-escape-by-gaining-control-of-a-user-s-computer-11679139203888.html
  6. is GPT shitfucked for anyone else

    https://www.tweaktown.com/news/90803/openai-forced-to-shut-down-chatgpt-following-bug-that-exposed-user-chat-history/index.html

  7. Alpaca 30b

  8. Originally posted by Stupid noob opened the secret cow portal Alpaca 30b


    Don't believe everything you read on the interwebs
  9. Um and AI could most definitely perceive it's world through sensory feedback, not just text. AI FAIL
  10. Meikai Heck This Schlong
    Originally posted by Jiggaboo_Johnson Um and AI could most definitely perceive it's world through sensory feedback, not just text. AI FAIL

    the only feedback a LLM has is textual. you can't get sensory feedback from eyes you don't have.
  11. Originally posted by Meikai the only feedback a LLM has is textual. you can't get sensory feedback from eyes you don't have.

    0s and 1s arent texts.
  12. Originally posted by Meikai the only feedback a LLM has is textual. you can't get sensory feedback from eyes you don't have.

    Um various sensors could easily be added...any intelligence would be able to figure that out. My fooking Roombas have sensory feedback.

    Imagine Einstein saying "oh I can't write relativity theory because I don't have a pen"
  13. Meikai Heck This Schlong
    Originally posted by Jiggaboo_Johnson Um various sensors could easily be added…any intelligence would be able to figure that out. My fooking Roombas have sensory feedback.

    Imagine Einstein saying "oh I can't write relativity theory because I don't have a pen"

    I mean, he literally couldn't write the theory of relativity without a pen, Jigs. Sure you could give him a pen, but if doesn't have one he doesn't have one.
  14. Meikai Heck This Schlong
    But yes you could use the same techniques behind a language model to build a neural network with an understanding of the statistical relationships between objects in a video feed, patterns in soundwaves, etc etc and have an increasingly more 'human' intelligence. But the point is, by building statistical relationships between words, it may be possible to create a rudimentary consciousness who understands existence in a way that's fundamentally dissimilar from the human experience.
  15. Originally posted by Jiggaboo_Johnson Don't believe everything you read on the interwebs

    okay

    Originally posted by Jiggaboo_Johnson Um and AI could most definitely perceive it's world through sensory feedback, not just text. AI FAIL



    Originally posted by Jiggaboo_Johnson Um various sensors could easily be added…any intelligence would be able to figure that out. My fooking Roombas have sensory feedback.

    Imagine Einstein saying "oh I can't write relativity theory because I don't have a pen"

    I dn't believe thjat
  16. Originally posted by Meikai I mean, he literally couldn't write the theory of relativity without a pen, Jigs. Sure you could give him a pen, but if doesn't have one he doesn't have one.

    Yes...but he'd have the intelligence to get a pen...that's the point here.

    He wouldn't respond when asked if he could write it..."no I don't have a pen"...unless he was an early troller..

    The AI should have responded "Well currently I lack the sensors but if installed I'd be able to used sensory feedback too"...if it was AI worth a shit of course.
  17. Originally posted by Jiggaboo_Johnson The AI should have responded

    You don't understand the purpose of this LLM then because it's not supposed to "simulate intelligence"

    It's not supposed to do what you think it should do. It does what it does because that's how it's programmed/prompted. It knows that it's a chatbot, the person in that example is essentially roleplaying with it

    Those responses are half hallucinatory anyways

    Originally posted by Jiggaboo_Johnson The AI should have responded "Well currently I lack the sensors but if installed I'd be able to used sensory feedback too"…if it was AI worth a shit of course.

    AI response to Your response to the AI response

    Both responses could be considered reasonable, depending on the context and the specific capabilities of the AI in question.

    The AI's response, "language models like myself can only process information through text-based sources," is an accurate statement for language models like OpenAI's GPT series or Google's BERT that are trained on large text corpora and designed to process natural language inputs in textual form. These language models lack the sensory input capabilities that humans possess, and as such, cannot "see" or "hear" in the same way that we do.

    On the other hand, Jigaboo Johnsons's response that the AI should have responded, "Well currently I lack the sensors but if installed I'd be able to use sensory feedback too," assumes that the AI in question is capable of processing sensory input if it were available. If the AI is indeed designed to process sensory inputs, then this response would be reasonable.

    In general, it's important to keep in mind that the capabilities and limitations of AI are determined by its design and the data it has been trained on. Different AI models may have different strengths and weaknesses, and may be designed for different purposes. So the "correct" response for an AI in a given situation will depend on the context and the specific AI model being used.
  18. if you installed eyes and ears VISUAL and SOUND input how would it know how to interpret that data without training it beforehand?

    This is a LANGUAGE model

    In general, it's important to keep in mind that the capabilities and limitations of AI are determined by its design and the data it has been trained on. Different AI models may have different strengths and weaknesses, and may be designed for different purposes. So the "correct" response for an AI in a given situation will depend on the context and the specific AI model being used.

    ^ what he said
  19. GPT4 actually can interpret visual data

    The following users say it would be alright if the author of this post didn't die in a fire!
Jump to Top