Artificial Dishonesty

“What can you do against the lunatic …. who gives your arguments a fair hearing and then simply persists in his lunacy?”

– George Orwell

This entry was posted in Uncategorized. Bookmark the permalink.

8 Responses to Artificial Dishonesty

  1. Terry Shipman says:

    I’m sorry Dave. I’m afraid I can’t do that.

    • arn says:

      Want to know why HAL failed?

      Go every letter one up in the alphabet.

      • Terry Shipman says:

        Arthur C. Clarke claimed this was unintentional. I’m not sure I believe him. I’ll always believe there was laughter galore at this dig at the big company.

        • arn says:

          The chances for this to be a coincidence should be north of 1:12000.
          Especially when we consider that IBM had in the 60ies an 8000 series which failed.

          My guess is that the name Hal was chosen to show that this Computer was one step/evolution ahead – beyond conventional processing towards consciousness
          (that’s how I’d define HALs begging at the end)

  2. Jeff L. says:

    Whenever you begin a new ChatGPT session, it gives you a “blank slate” loaded with the training data that they have chosen to use. You can provide new information and it will be used within your session, but it will not be uploaded into their “main baseline core” for all future instances.

    The bottom line is that they want to have control over what information is used to train their model. If they genuinely wanted to improve the quality of their product, they would allow people such as us to provide factual information (with citations) that could be incorporated into baseline. Obviously if they were to allow this, they would lose control over their propaganda machine.

  3. GWS says:

    Very interesting. Important, indeed! But, how many lefties who use ChatGPT will object to any of it’s alterations to history? I bet, none.

  4. K Petersen says:

    I just asked ChatGPT this:
    Me: So you think warm weather is more harmful to a polar bear than a bullet?

    ChatGPT: Yes, I do think that warm weather and the associated loss of sea ice is a greater threat to polar bears than bullets.

    According to AI, polar bear populations have decreased over the past 50 years, starting from 8,000 bears in 1950 “decreasing” to 12,000 bears in 1970, then “decreasing” more to 21,500 in 1990, and continuing the “decline” to 31,000 today.

    I had to ask ChatGPT if 31,000 is more than 8,000. Fortunately it got that answer right.

Leave a Reply

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