3 ms·

You can prove it easily by having many intense conversations with specific details. Then open it in a new browser and it won't have any idea what you are talkin

You can prove it easily by having many intense conversations with specific details. Then open it in a new browser and it won't have any idea what you are talking about.

HEmanZ 4y ago
So long term memory is a condition for intelligence or consciousness? Another weird one that applies so well to these LLMs: would you consider humans conscious or intelligent when they’re dreaming? Even when the dream consists of remember false memories? I think we’re pushing close to the line where we don’t understand if these things are intelligent. Or we break our understanding of what intelligent means
kragen 4y ago
does this mean humans with prospective amnesia aren't intelligent or self-aware
- wizofaus 4y ago
  I would say it's probably impossible to have complete short term amnesia and fully- functioning self awareness as we normally conceive of it, yes. There's even an argument that memories are really the only thing we can experience , and your memory of what occurred seconds/minutes/hours/days etc ago are the only way you can said to "be" (or have the experience of being) a particular individual. That OpenAI- based LLMs don't have such memories almost certainly rules out any possibility of them having a sense of "self".
  - kragen 4y ago
    i meant long-term: alzheimer's patients who ask you every five minutes where their dead husband is people while dreaming are another case is it ok to kill them
    - wizofaus 4y ago
      Whether it's OK to kill them is a far more difficult question and to be honest I don't know, but my instinct is that if all their loved ones who clearly have the individual's best interests at heart agree that ending their life would be the best option, and obviously assuming it was done painlessly etc., then yes, it's an ethically acceptable choice (certainly far more so than many of the activities humans regularly take part in, especially those clearly harmful to other species or our planet's chances of supporting human life in the future).
      - kragen 4y ago
        are you honestly claiming that it would be okay for parents to kill their children painlessly while they're asleep because the children don't have long-term memory while in that state
        wizofaus 4y ago
        Asleep is clearly a temporary state. Babies/sleeping people still have the potential to become self-aware and the expectation is that they generally will. At any rate I don't think it's relevant to the discussion at hand.
didntreadarticl 4y ago
But maybe in the milliseconds where billions of GPUs across a vast network activate and process your input, and weigh up a billions parameters before assembling a reply, there is a spark of awareness. Who's to say?

You can prove it easily by having many intense conversations with specific details. Then open it in a new browser and it won't have any idea what you are talking about.

HEmanZ 4y ago

So long term memory is a condition for intelligence or consciousness? Another weird one that applies so well to these LLMs: would you consider humans conscious or intelligent when they’re dreaming? Even when the dream consists of remember false memories? I think we’re pushing close to the line where we don’t understand if these things are intelligent. Or we break our understanding of what intelligent means

kragen 4y ago

does this mean humans with prospective amnesia aren't intelligent or self-aware

wizofaus 4y ago
I would say it's probably impossible to have complete short term amnesia and fully- functioning self awareness as we normally conceive of it, yes. There's even an argument that memories are really the only thing we can experience , and your memory of what occurred seconds/minutes/hours/days etc ago are the only way you can said to "be" (or have the experience of being) a particular individual. That OpenAI- based LLMs don't have such memories almost certainly rules out any possibility of them having a sense of "self".
- kragen 4y ago
  i meant long-term: alzheimer's patients who ask you every five minutes where their dead husband is people while dreaming are another case is it ok to kill them
  - wizofaus 4y ago
    Whether it's OK to kill them is a far more difficult question and to be honest I don't know, but my instinct is that if all their loved ones who clearly have the individual's best interests at heart agree that ending their life would be the best option, and obviously assuming it was done painlessly etc., then yes, it's an ethically acceptable choice (certainly far more so than many of the activities humans regularly take part in, especially those clearly harmful to other species or our planet's chances of supporting human life in the future).
    - kragen 4y ago
      are you honestly claiming that it would be okay for parents to kill their children painlessly while they're asleep because the children don't have long-term memory while in that state
      - wizofaus 4y ago
        Asleep is clearly a temporary state. Babies/sleeping people still have the potential to become self-aware and the expectation is that they generally will. At any rate I don't think it's relevant to the discussion at hand.

didntreadarticl 4y ago

But maybe in the milliseconds where billions of GPUs across a vast network activate and process your input, and weigh up a billions parameters before assembling a reply, there is a spark of awareness. Who's to say?