Comments by "David" (@stoogel) on "TheAIGRID" channel.

  1. Please do not ascribe traits of a sentient being to the clever statistical model. There's been enough panic about that from people who don't know what they're talking about.
    2
  2.  @하하호호-h3u  It can do certain things humans do with prompting. It has no agency, no mental states that respond to external stimuli, no autonomy. At its core it's a statistical model with the ability to guess responses based on pre-trained weights. The "human" traits you claim it has are the result of it being programmed to feed its prompts and output back into itself. There are no "papers and reports" of any repute that would claim it's anything more.
    2
  3. Because it's not capable of real logic and reasoning.
    2
  4. Now that there's not only profit motive but a geopolitical aspect, it's obvious nothing will slow down. We're diving head-first into either our ruin or ascendance as a species. Government agencies are pushing for further development, but I wonder if public policy will ever catch up in time to prepare us for the economic and eventual existential consequences.
    1
  5.  @darkandtwisted  In a capitalist society, everyone should fear AI. When your ability to survive is tied to the value of your labor, a technology that decimates the value of all labor will negatively affect your survival.
    1
  6. Governments need to prepare for the economic consequences now, because the only thing slowing this down is the sheer amount of computing power required.
    1
  7. Microsoft? Noble? Cautious? High moral standards? Microsoft?
    1
  8. Part of the advancement of these models is the ability to get better results from the same amount of training data or less.
    1
  9. You're ready to trust something that calls itself "TruthGPT" to be unbiased?
    1