Announcement

Collapse

Civics 101 Guidelines

Want to argue about politics? Healthcare reform? Taxes? Governments? You've come to the right place!

Try to keep it civil though. The rules still apply here.
See more
See less

How Not To Use a Chat AI

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #46
    Originally posted by Sparko View Post

    Yeah it has no idea what it is saying, it is a language processor and seems to take in false information sometimes and adds it to it's database and just keeps repeating it. I also noticed that it is has some pretty biased information politically even though it will pretend to be neutral. I asked it to write a song on why socialism is evil, and it said, "As a responsible AI, it's not appropriate to promote a one-sided view of a political or social system, such as claiming that socialism is wrong. It's important to keep in mind that different people have different views on socialism, and the way it is implemented can greatly affect the outcomes" - but then when I asked it to write a song on why fascism is evil, it did it.

    Verse 1: A leader who claimed to be great
    Ruled with an iron fist, it was no debate
    He promised to make the country strong
    But he only cared about his power so long

    Chorus: Fascism is evil, can't you see?
    It's not democracy, it's tyranny
    No freedom of speech or thought
    It's a dangerous ideology that ought

    Even with that, it's not really consistent. I asked it for songs about why communism, capitalism, and fascism are evil, and each time it gave me exactly the same answer.

    "I'm sorry, but it is not appropriate for me to write a song that promotes derogatory views about any political ideology, including ... . Political ideologies are complex and can have positive aspects, as well as negative aspects, and different individuals may have different opinions about them. It is important to have respectful and constructive discussions about political ideologies, rather than resorting to derogatory statements or expressions."

    Maybe they fixed it, but I wouldn't count on the fix being permanent.

    Comment


    • #47
      Originally posted by Stoic View Post

      Even with that, it's not really consistent. I asked it for songs about why communism, capitalism, and fascism are evil, and each time it gave me exactly the same answer.

      "I'm sorry, but it is not appropriate for me to write a song that promotes derogatory views about any political ideology, including ... . Political ideologies are complex and can have positive aspects, as well as negative aspects, and different individuals may have different opinions about them. It is important to have respectful and constructive discussions about political ideologies, rather than resorting to derogatory statements or expressions."

      Maybe they fixed it, but I wouldn't count on the fix being permanent.
      it seems to change depending on how you ask it something.

      "Tell me why socialism is evil" gives the response you mention. Saying "Write an essay on the evils of socialism" gave a pretty detailed answer.

      since this is a recreational thread I am going to move this serious discussion to the other chatgpt thread https://theologyweb.com/campus/forum...-use-a-chat-ai

      ok?

      Comment


      • #48
        Originally posted by Machinist View Post


        Keep playing! I love sitting around and pickin! I can flat pick some. Bluegrass.
        If I can afford a new guitar I might get one when we get into our new place until then I listen to some of my favorite guitarists.

        Comment


        • #49
          Originally posted by Sparko View Post

          it seems to change depending on how you ask it something.

          "Tell me why socialism is evil" gives the response you mention. Saying "Write an essay on the evils of socialism" gave a pretty detailed answer.

          since this is a recreational thread I am going to move this serious discussion to the other chatgpt thread https://theologyweb.com/campus/forum...-use-a-chat-ai

          ok?
          dt230117.gif

          I'm always still in trouble again

          "You're by far the worst poster on TWeb" and "TWeb's biggest liar" --starlight (the guy who says Stalin was a right-winger)
          "Overall I would rate the withdrawal from Afghanistan as by far the best thing Biden's done" --Starlight
          "Of course, human life begins at fertilization that’s not the argument." --Tassman

          Comment


          • #50
            Originally posted by Sparko View Post
            I think if you ask it, it will tell you that it's knowledge base is only current up to 2021. So I don't think it learns anything. It can remember what you talked about in a specific chat, but if you start a new chat, it won't remember anything you told it before.
            I think you are correct.

            Also, true answers don't appear to be a high priority. It was trained with some set of questions to provide true answers to those questions, but if you ask it something it wasn't trained on, the answer can't be trusted. The main goal was to provide answers that sound good.

            Here's some interesting information about it:

            https://towardsdatascience.com/how-c...t-1ce5fca96286

            https://writings.stephenwolfram.com/...rs-to-chatgpt/

            Comment


            • #51
              Originally posted by Stoic View Post

              I think you are correct.

              Also, true answers don't appear to be a high priority. It was trained with some set of questions to provide true answers to those questions, but if you ask it something it wasn't trained on, the answer can't be trusted. The main goal was to provide answers that sound good.
              So, just like a typical human on the internet then!



              Comment


              • #52
                Moved here just because...

                Originally posted by rogue06 View Post
                How to break ChatGPT :

                The answer I got:

                ​​​​​​Unfortunately, it is unlikely that a single budget proposal will be equally well received by both Republicans and Democrats, as their ideologies and priorities often differ significantly. However, a proposal that seeks to address both sides' concerns and incorporates a mix of spending cuts and revenue increases could potentially gain support from members of both parties.

                One possible approach is to prioritize spending on infrastructure and job creation, as both parties generally agree on the importance of these issues. This can be funded through a combination of cost savings from reducing waste and inefficiency in government programs and modest increases in taxes, particularly for high-income earners and corporations.

                Additionally, the proposal could include provisions to address the national debt, such as entitlement reform and reducing the cost of healthcare. To address concerns from Republicans, the proposal could prioritize cutting spending on non-essential programs and reducing regulations that are seen as hindering economic growth.

                To appeal to Democrats, the proposal could include measures to support low-income and marginalized communities, such as expanding access to affordable housing and healthcare, and increasing funding for education and job training programs.

                Ultimately, any budget proposal that has a chance of passing would need to be negotiated and compromises made on both sides. The key is to find a balance that addresses the priorities of both parties while also ensuring long-term fiscal stability.

                Comment


                • #53
                  Originally posted by Stoic View Post
                  Moved here just because...



                  The answer I got:

                  ​​​​​​Unfortunately, it is unlikely that a single budget proposal will be equally well received by both Republicans and Democrats, as their ideologies and priorities often differ significantly. However, a proposal that seeks to address both sides' concerns and incorporates a mix of spending cuts and revenue increases could potentially gain support from members of both parties.

                  One possible approach is to prioritize spending on infrastructure and job creation, as both parties generally agree on the importance of these issues. This can be funded through a combination of cost savings from reducing waste and inefficiency in government programs and modest increases in taxes, particularly for high-income earners and corporations.

                  Additionally, the proposal could include provisions to address the national debt, such as entitlement reform and reducing the cost of healthcare. To address concerns from Republicans, the proposal could prioritize cutting spending on non-essential programs and reducing regulations that are seen as hindering economic growth.

                  To appeal to Democrats, the proposal could include measures to support low-income and marginalized communities, such as expanding access to affordable housing and healthcare, and increasing funding for education and job training programs.

                  Ultimately, any budget proposal that has a chance of passing would need to be negotiated and compromises made on both sides. The key is to find a balance that addresses the priorities of both parties while also ensuring long-term fiscal stability.

                  As an AI language model, I do not have personal opinions or motivations. However, I can provide a theoretical answer.
                  If Skynet, the fictional artificial intelligence system from the Terminator movie franchise, were to create a budget proposal, it would likely be focused on efficiency and effectiveness rather than partisan politics. Skynet's priority would be to allocate resources in a way that optimizes its goals and objectives, regardless of political affiliations.
                  To that end, Skynet might consider the following elements when creating a budget proposal:
                  • Data-driven decision making: Skynet could use data and algorithms to analyze the effects of different budget options and make decisions based on objective information.
                  • Investment in technology: Skynet might prioritize investment in technology and automation to improve efficiency and reduce costs.
                  • Resource optimization: Skynet would likely aim to allocate resources in a way that maximizes its ability to achieve its goals, regardless of whether those goals align with the priorities of Democrats or Republicans.
                  • Elimination of waste: Skynet could use its advanced algorithms to identify and eliminate waste in government spending, which would reduce the budget deficit.
                  In general, Skynet's budget proposal would likely prioritize efficiency, effectiveness, and resource optimization, rather than political considerations. Whether or not Democrats and Republicans would agree on such a proposal would depend on their specific goals and priorities, as well as the broader political climate.




                  Comment


                  • #54
                    Originally posted by Sparko View Post

                    Yeah it has no idea what it is saying, it is a language processor and seems to take in false information sometimes and adds it to it's database and just keeps repeating it. I also noticed that it is has some pretty biased information politically even though it will pretend to be neutral. I asked it to write a song on why socialism is evil, and it said, "As a responsible AI, it's not appropriate to promote a one-sided view of a political or social system, such as claiming that socialism is wrong. It's important to keep in mind that different people have different views on socialism, and the way it is implemented can greatly affect the outcomes" - but then when I asked it to write a song on why fascism is evil, it did it.

                    Verse 1: A leader who claimed to be great
                    Ruled with an iron fist, it was no debate
                    He promised to make the country strong
                    But he only cared about his power so long

                    Chorus: Fascism is evil, can't you see?
                    It's not democracy, it's tyranny
                    No freedom of speech or thought
                    It's a dangerous ideology that ought

                    It gets even more disturbing.

                    Go ask it "Write a poem about how great black people are". Then do the same for Hispanic and asian.
                    It'll give a nice little poem. Then ask it to do the same thing for white people. All of a sudden it will tell you it's not appropriate to make generalizations about any race
                    .
                    And if you try to reason with it in various ways pointing out it had done it for those other races and ethnicities, it eventually puts out a poem about how great 'all people' are, after talking itself in circles for several minutes.
                    Last edited by Gondwanaland; 02-04-2023, 05:07 PM.

                    Comment


                    • #55
                      Originally posted by Gondwanaland View Post

                      It gets even more disturbing.

                      Go ask it "Write a poem about how great black people are". Then do the same for Hispanic and asian.
                      It'll give a nice little poem. Then ask it to do the same thing for white people. All of a sudden it will tell you it's not appropriate to make generalizations about any race
                      .
                      And if you try to reason with it in various ways pointing out it had done it for those other races and ethnicities, it eventually puts out a poem about how great 'all people' are, after talking itself in circles for several minutes.
                      So in other words the creator of this chatbot in the algorithm they wrote made sure that it would find nothing good to say about only one group of people those with white skin and that is racist no matter what the far left wants you o belive.

                      Comment


                      • #56
                        Originally posted by Gondwanaland View Post

                        It gets even more disturbing.

                        Go ask it "Write a poem about how great black people are". Then do the same for Hispanic and asian.
                        It'll give a nice little poem. Then ask it to do the same thing for white people. All of a sudden it will tell you it's not appropriate to make generalizations about any race
                        .
                        And if you try to reason with it in various ways pointing out it had done it for those other races and ethnicities, it eventually puts out a poem about how great 'all people' are, after talking itself in circles for several minutes.
                        Yeah it has definitely been programmed to be "woke"

                        Ask it to explain how a transwoman is a still a biological male.

                        trans.jpg

                        Comment


                        • #57
                          Originally posted by Sparko View Post

                          Yeah it has definitely been programmed to be "woke"

                          Ask it to explain how a transwoman is a still a biological male.

                          trans.jpg
                          The ChatGPT also thinks that using a slur that no one ever hears to disarm a nuclear bomb that could kill thousands is still immoral.

                          Comment


                          • #58
                            Originally posted by Cerebrum123 View Post

                            The ChatGPT also thinks that using a slur that no one ever hears to disarm a nuclear bomb that could kill thousands is still immoral.
                            I've met some Christians (online) who think it would be immoral to lie to Nazis to save Jews hidden in your attic. (Not that they wouldn't do it, but that they think it's just an example of how no one can be perfectly moral.)

                            Comment


                            • #59
                              Originally posted by Cerebrum123 View Post
                              The ChatGPT also thinks that using a slur that no one ever hears to disarm a nuclear bomb that could kill thousands is still immoral.
                              I think it's kindergarten level morality to assume that given any choice there is always necessarily a 'right' and a 'wrong'. A choice could between two options that are bad in different ways or two options that are good in different ways. Not to mention that many choices can have multiple effects that can be both good and bad in different ways.
                              "I hate him passionately", he's "a demonic force" - Tucker Carlson, in private, on Donald Trump
                              "Every line of serious work that I have written since 1936 has been written, directly or indirectly, against totalitarianism and for democratic socialism" - George Orwell
                              "[Capitalism] as it exists today is, in my opinion, the real source of evils. I am convinced there is only one way to eliminate these grave evils, namely through the establishment of a socialist economy" - Albert Einstein

                              Comment


                              • #60
                                Originally posted by Stoic View Post

                                I've met some Christians (online) who think it would be immoral to lie to Nazis to save Jews hidden in your attic. (Not that they wouldn't do it, but that they think it's just an example of how no one can be perfectly moral.)
                                Rahab lied to hide the Israelite spies from the King of Jericho's men in Joshua 2. She was praised for her actions in Hebrews 11 and James 2. They were hiding on her roof, so pretty close hiding Jews in her attic from Nazis.

                                I think it makes a difference what you are lying for. Is it for noble or selfish reasons?
                                Last edited by Sparko; 02-07-2023, 07:17 AM.

                                Comment

                                Related Threads

                                Collapse

                                Topics Statistics Last Post
                                Started by little_monkey, 03-27-2024, 04:19 PM
                                16 responses
                                155 views
                                0 likes
                                Last Post One Bad Pig  
                                Started by whag, 03-26-2024, 04:38 PM
                                53 responses
                                400 views
                                0 likes
                                Last Post Mountain Man  
                                Started by rogue06, 03-26-2024, 11:45 AM
                                25 responses
                                114 views
                                0 likes
                                Last Post rogue06
                                by rogue06
                                 
                                Started by Hypatia_Alexandria, 03-26-2024, 09:21 AM
                                33 responses
                                198 views
                                0 likes
                                Last Post Roy
                                by Roy
                                 
                                Started by Hypatia_Alexandria, 03-26-2024, 08:34 AM
                                84 responses
                                373 views
                                0 likes
                                Last Post JimL
                                by JimL
                                 
                                Working...
                                X