Skip to content
  • Categories
  • Recent
  • Tags
  • Popular
  • World
  • Users
  • Groups
Skins
  • Light
  • Brite
  • Cerulean
  • Cosmo
  • Flatly
  • Journal
  • Litera
  • Lumen
  • Lux
  • Materia
  • Minty
  • Morph
  • Pulse
  • Sandstone
  • Simplex
  • Sketchy
  • Spacelab
  • United
  • Yeti
  • Zephyr
  • Dark
  • Cyborg
  • Darkly
  • Quartz
  • Slate
  • Solar
  • Superhero
  • Vapor

  • Default (Cyborg)
  • No Skin
Collapse
Brand Logo

CIRCLE WITH A DOT

  1. Home
  2. Uncategorized
  3. I am asking people once again to stop describing #AI using human words and attributes.

I am asking people once again to stop describing #AI using human words and attributes.

Scheduled Pinned Locked Moved Uncategorized
47 Posts 35 Posters 25 Views
  • Oldest to Newest
  • Newest to Oldest
  • Most Votes
Reply
  • Reply as topic
Log in to reply
This topic has been deleted. Only users with topic management privileges can see it.
  • drahardja@sfba.socialD drahardja@sfba.social

    I am asking people once again to stop describing #AI using human words and attributes.

    AI doesn’t answer—it regurgitates plausible output that happens to be right.
    AI doesn’t hallucinate—it regurgitates plausible output that happens to be wrong.
    AI doesn’t have agency—it regurgitates plausible output that is consumed by a computer program.
    AI doesn’t think—it regurgitates plausible output that is fed back into itself to try again.
    AI doesn’t create—it regurgitates a mixture of its training data.

    Remember: the robot is not a person.

    pattykimura@beige.partyP This user is from outside of this forum
    pattykimura@beige.partyP This user is from outside of this forum
    pattykimura@beige.party
    wrote last edited by
    #29

    @drahardja

    💯

    The most resonate description for me is by Dr Emily Bender who compares it to a Magic Eight Ball, a 64 bit Magic Eight Ball.

    I remain shocked at how easily, relatively "smart" people embrace AI as a cure-all. How CEOs become cheerleaders to subordinates of an unproven robot to "extrude" managerial "decisions" without question.

    Crazy.

    1 Reply Last reply
    0
    • lehtimaeki@snapp.socialL lehtimaeki@snapp.social

      @drahardja
      Also, using "AI" for LLMs is misleading. There's no I.

      mrgrumpymonkey@mastodon.socialM This user is from outside of this forum
      mrgrumpymonkey@mastodon.socialM This user is from outside of this forum
      mrgrumpymonkey@mastodon.social
      wrote last edited by
      #30

      @lehtimaeki @drahardja There's no A either.

      forthy42@mastodon.net2o.deF 1 Reply Last reply
      0
      • drahardja@sfba.socialD drahardja@sfba.social

        I am asking people once again to stop describing #AI using human words and attributes.

        AI doesn’t answer—it regurgitates plausible output that happens to be right.
        AI doesn’t hallucinate—it regurgitates plausible output that happens to be wrong.
        AI doesn’t have agency—it regurgitates plausible output that is consumed by a computer program.
        AI doesn’t think—it regurgitates plausible output that is fed back into itself to try again.
        AI doesn’t create—it regurgitates a mixture of its training data.

        Remember: the robot is not a person.

        rensbloom@zug.networkR This user is from outside of this forum
        rensbloom@zug.networkR This user is from outside of this forum
        rensbloom@zug.network
        wrote last edited by
        #31

        @drahardja The worst thing I have heard was a statement "you know who can dothis very well? Copilot!"

        1 Reply Last reply
        0
        • drahardja@sfba.socialD drahardja@sfba.social

          I am asking people once again to stop describing #AI using human words and attributes.

          AI doesn’t answer—it regurgitates plausible output that happens to be right.
          AI doesn’t hallucinate—it regurgitates plausible output that happens to be wrong.
          AI doesn’t have agency—it regurgitates plausible output that is consumed by a computer program.
          AI doesn’t think—it regurgitates plausible output that is fed back into itself to try again.
          AI doesn’t create—it regurgitates a mixture of its training data.

          Remember: the robot is not a person.

          chriscorrigan@cosocial.caC This user is from outside of this forum
          chriscorrigan@cosocial.caC This user is from outside of this forum
          chriscorrigan@cosocial.ca
          wrote last edited by
          #32

          @drahardja it’s a magic 8 ball that tells you what you want to hear.

          1 Reply Last reply
          0
          • drahardja@sfba.socialD drahardja@sfba.social

            I am asking people once again to stop describing #AI using human words and attributes.

            AI doesn’t answer—it regurgitates plausible output that happens to be right.
            AI doesn’t hallucinate—it regurgitates plausible output that happens to be wrong.
            AI doesn’t have agency—it regurgitates plausible output that is consumed by a computer program.
            AI doesn’t think—it regurgitates plausible output that is fed back into itself to try again.
            AI doesn’t create—it regurgitates a mixture of its training data.

            Remember: the robot is not a person.

            yala@degrowth.socialY This user is from outside of this forum
            yala@degrowth.socialY This user is from outside of this forum
            yala@degrowth.social
            wrote last edited by
            #33

            @drahardja
            AI is not trained.

            The preproduction of the model is also included in the anthropomorphisation. Next they will be speaking of giving birth to new models.

            @harold

            drahardja@sfba.socialD 1 Reply Last reply
            0
            • yala@degrowth.socialY yala@degrowth.social

              @drahardja
              AI is not trained.

              The preproduction of the model is also included in the anthropomorphisation. Next they will be speaking of giving birth to new models.

              @harold

              drahardja@sfba.socialD This user is from outside of this forum
              drahardja@sfba.socialD This user is from outside of this forum
              drahardja@sfba.social
              wrote last edited by
              #34

              @yala @harold What’s a better term?

              dallo@pouet.chapril.orgD 1 Reply Last reply
              0
              • drahardja@sfba.socialD drahardja@sfba.social

                I am asking people once again to stop describing #AI using human words and attributes.

                AI doesn’t answer—it regurgitates plausible output that happens to be right.
                AI doesn’t hallucinate—it regurgitates plausible output that happens to be wrong.
                AI doesn’t have agency—it regurgitates plausible output that is consumed by a computer program.
                AI doesn’t think—it regurgitates plausible output that is fed back into itself to try again.
                AI doesn’t create—it regurgitates a mixture of its training data.

                Remember: the robot is not a person.

                standmit@mastodon.luanti.ruS This user is from outside of this forum
                standmit@mastodon.luanti.ruS This user is from outside of this forum
                standmit@mastodon.luanti.ru
                wrote last edited by
                #35

                @drahardja
                Replace the word "AI" with "human" in that statement and it would still be true.

                1 Reply Last reply
                0
                • drahardja@sfba.socialD drahardja@sfba.social

                  @yala @harold What’s a better term?

                  dallo@pouet.chapril.orgD This user is from outside of this forum
                  dallo@pouet.chapril.orgD This user is from outside of this forum
                  dallo@pouet.chapril.org
                  wrote last edited by
                  #36

                  @drahardja
                  assimilated ? categorized ? grouped ? patterned ?

                  @yala @harold

                  1 Reply Last reply
                  0
                  • drahardja@sfba.socialD drahardja@sfba.social

                    I am asking people once again to stop describing #AI using human words and attributes.

                    AI doesn’t answer—it regurgitates plausible output that happens to be right.
                    AI doesn’t hallucinate—it regurgitates plausible output that happens to be wrong.
                    AI doesn’t have agency—it regurgitates plausible output that is consumed by a computer program.
                    AI doesn’t think—it regurgitates plausible output that is fed back into itself to try again.
                    AI doesn’t create—it regurgitates a mixture of its training data.

                    Remember: the robot is not a person.

                    paniczgodek@functional.cafeP This user is from outside of this forum
                    paniczgodek@functional.cafeP This user is from outside of this forum
                    paniczgodek@functional.cafe
                    wrote last edited by
                    #37

                    @drahardja we describe various phenomena using human words and attributes because we are human

                    (And computers were thinking at least since someone implemented an hourglass icon in them)

                    1 Reply Last reply
                    0
                    • drahardja@sfba.socialD drahardja@sfba.social

                      I am asking people once again to stop describing #AI using human words and attributes.

                      AI doesn’t answer—it regurgitates plausible output that happens to be right.
                      AI doesn’t hallucinate—it regurgitates plausible output that happens to be wrong.
                      AI doesn’t have agency—it regurgitates plausible output that is consumed by a computer program.
                      AI doesn’t think—it regurgitates plausible output that is fed back into itself to try again.
                      AI doesn’t create—it regurgitates a mixture of its training data.

                      Remember: the robot is not a person.

                      unchartedworlds@scicomm.xyzU This user is from outside of this forum
                      unchartedworlds@scicomm.xyzU This user is from outside of this forum
                      unchartedworlds@scicomm.xyz
                      wrote last edited by
                      #38

                      @drahardja

                      There's also the option of adding quotes around "AI", as a way to indicate "so-called, not what it actually is".

                      1 Reply Last reply
                      0
                      • drahardja@sfba.socialD drahardja@sfba.social

                        I am asking people once again to stop describing #AI using human words and attributes.

                        AI doesn’t answer—it regurgitates plausible output that happens to be right.
                        AI doesn’t hallucinate—it regurgitates plausible output that happens to be wrong.
                        AI doesn’t have agency—it regurgitates plausible output that is consumed by a computer program.
                        AI doesn’t think—it regurgitates plausible output that is fed back into itself to try again.
                        AI doesn’t create—it regurgitates a mixture of its training data.

                        Remember: the robot is not a person.

                        mlevison@hachyderm.ioM This user is from outside of this forum
                        mlevison@hachyderm.ioM This user is from outside of this forum
                        mlevison@hachyderm.io
                        wrote last edited by
                        #39

                        @drahardja +1000

                        1 Reply Last reply
                        0
                        • drahardja@sfba.socialD drahardja@sfba.social

                          I am asking people once again to stop describing #AI using human words and attributes.

                          AI doesn’t answer—it regurgitates plausible output that happens to be right.
                          AI doesn’t hallucinate—it regurgitates plausible output that happens to be wrong.
                          AI doesn’t have agency—it regurgitates plausible output that is consumed by a computer program.
                          AI doesn’t think—it regurgitates plausible output that is fed back into itself to try again.
                          AI doesn’t create—it regurgitates a mixture of its training data.

                          Remember: the robot is not a person.

                          gunstick@mastodon.opencloud.luG This user is from outside of this forum
                          gunstick@mastodon.opencloud.luG This user is from outside of this forum
                          gunstick@mastodon.opencloud.lu
                          wrote last edited by
                          #40

                          @drahardja
                          I think bullshit machine is adequate.
                          Maybe not exactly nice towards the bull and shit in general. At least it's not a human condition (usually).

                          forthy42@mastodon.net2o.deF 1 Reply Last reply
                          0
                          • drahardja@sfba.socialD drahardja@sfba.social

                            I am asking people once again to stop describing #AI using human words and attributes.

                            AI doesn’t answer—it regurgitates plausible output that happens to be right.
                            AI doesn’t hallucinate—it regurgitates plausible output that happens to be wrong.
                            AI doesn’t have agency—it regurgitates plausible output that is consumed by a computer program.
                            AI doesn’t think—it regurgitates plausible output that is fed back into itself to try again.
                            AI doesn’t create—it regurgitates a mixture of its training data.

                            Remember: the robot is not a person.

                            forthy42@mastodon.net2o.deF This user is from outside of this forum
                            forthy42@mastodon.net2o.deF This user is from outside of this forum
                            forthy42@mastodon.net2o.de
                            wrote last edited by
                            #41

                            @drahardja Given how less smart humans produce answers, we'd rather hope not to know with these details how their model works…

                            1 Reply Last reply
                            0
                            • gunstick@mastodon.opencloud.luG gunstick@mastodon.opencloud.lu

                              @drahardja
                              I think bullshit machine is adequate.
                              Maybe not exactly nice towards the bull and shit in general. At least it's not a human condition (usually).

                              forthy42@mastodon.net2o.deF This user is from outside of this forum
                              forthy42@mastodon.net2o.deF This user is from outside of this forum
                              forthy42@mastodon.net2o.de
                              wrote last edited by
                              #42

                              @gunstick @drahardja Bullshit machines would produce bullshit no matter what kind of input you give. The Autocompletion Parrot doesn't. If you manage to guide it to autocomplete a good conversation starter, you can somewhat get something not too far from what you could have written yourself.

                              gunstick@mastodon.opencloud.luG 1 Reply Last reply
                              0
                              • forthy42@mastodon.net2o.deF forthy42@mastodon.net2o.de

                                @gunstick @drahardja Bullshit machines would produce bullshit no matter what kind of input you give. The Autocompletion Parrot doesn't. If you manage to guide it to autocomplete a good conversation starter, you can somewhat get something not too far from what you could have written yourself.

                                gunstick@mastodon.opencloud.luG This user is from outside of this forum
                                gunstick@mastodon.opencloud.luG This user is from outside of this forum
                                gunstick@mastodon.opencloud.lu
                                wrote last edited by
                                #43

                                @forthy42 @drahardja oh yes indeed. User error. By the meme "every machine is a smoke machine if operated wrong enough" I present you the "every AI is a bullshit machine if prompted wrong enough"

                                forthy42@mastodon.net2o.deF 1 Reply Last reply
                                0
                                • mrgrumpymonkey@mastodon.socialM mrgrumpymonkey@mastodon.social

                                  @lehtimaeki @drahardja There's no A either.

                                  forthy42@mastodon.net2o.deF This user is from outside of this forum
                                  forthy42@mastodon.net2o.deF This user is from outside of this forum
                                  forthy42@mastodon.net2o.de
                                  wrote last edited by
                                  #44

                                  @mrgrumpymonkey @lehtimaeki @drahardja It actually is a compression function of human generated text together with a probability optimizer to decompress the most likely text.

                                  So yes, it's not intelligent, and it is not artificial. It is just another sloppy way to search the Internet, and find something that is close, but not actually there.

                                  1 Reply Last reply
                                  0
                                  • gunstick@mastodon.opencloud.luG gunstick@mastodon.opencloud.lu

                                    @forthy42 @drahardja oh yes indeed. User error. By the meme "every machine is a smoke machine if operated wrong enough" I present you the "every AI is a bullshit machine if prompted wrong enough"

                                    forthy42@mastodon.net2o.deF This user is from outside of this forum
                                    forthy42@mastodon.net2o.deF This user is from outside of this forum
                                    forthy42@mastodon.net2o.de
                                    wrote last edited by
                                    #45

                                    @gunstick @drahardja Sometimes it is hard to get an AI to accept the bullshit we humans created.

                                    I wanted Deepseek to summarize the Iran war Trump started. It told me this war doesn't exist. I gave it Web Sites from established and reliable sources to read. It said someone managed to inject a hoax there or that it is a satirical fiction. Nobody, not even Donald Trump would be that stupid to start this war!

                                    It was really hard work to get it to accept that our timeline is not only the one where Biff Tannen became president, but the one where Idiocracy happened already in 2025.

                                    It was really funny to see how the training to prevent bullshit output turned against the bullshit we have in our reality.

                                    1 Reply Last reply
                                    0
                                    • drahardja@sfba.socialD drahardja@sfba.social

                                      I am asking people once again to stop describing #AI using human words and attributes.

                                      AI doesn’t answer—it regurgitates plausible output that happens to be right.
                                      AI doesn’t hallucinate—it regurgitates plausible output that happens to be wrong.
                                      AI doesn’t have agency—it regurgitates plausible output that is consumed by a computer program.
                                      AI doesn’t think—it regurgitates plausible output that is fed back into itself to try again.
                                      AI doesn’t create—it regurgitates a mixture of its training data.

                                      Remember: the robot is not a person.

                                      strutsulf@mastodon.socialS This user is from outside of this forum
                                      strutsulf@mastodon.socialS This user is from outside of this forum
                                      strutsulf@mastodon.social
                                      wrote last edited by
                                      #46

                                      @drahardja Chat bots (etc.) only "say"and "claim" things, "give answers", "tell" you things, "advice" you, and so on, in the same sense that a textbook or any other text "says" or "claims" things etc. And as in the case of other texts it is the author or publisher who, in the last analysis, is responsible for everything said or claimed etc. in the text, so with "AI" generated text. What we should do is hold tech giants responsible for what is said etc in AI generated text. They're the "authors".

                                      1 Reply Last reply
                                      0
                                      • robin_wren@mastoart.socialR robin_wren@mastoart.social

                                        @jairajdevadiga @drahardja I think you know what I meant, though.

                                        prinlu@0x.trans.failP This user is from outside of this forum
                                        prinlu@0x.trans.failP This user is from outside of this forum
                                        prinlu@0x.trans.fail
                                        wrote last edited by
                                        #47

                                        @robin_wren @jairajdevadiga @drahardja it makes me sick every time when someone goes "nature invented this or that feature of a living otgsnism"

                                        1 Reply Last reply
                                        0
                                        Reply
                                        • Reply as topic
                                        Log in to reply
                                        • Oldest to Newest
                                        • Newest to Oldest
                                        • Most Votes


                                        • Login

                                        • Login or register to search.
                                        • First post
                                          Last post
                                        0
                                        • Categories
                                        • Recent
                                        • Tags
                                        • Popular
                                        • World
                                        • Users
                                        • Groups