• 013zen
    164
    For the past some odd months, I've been working on a side project which utilizes A.I. to map out beliefs and tensions from user inputs, and respond in a manner meant to help the user reflect and engage within their beliefs, in an environment neither meant for spectacle nor coersion (in the sense of being told what to believe).

    I started this project because, I tried using chat GPT to discuss philosophy, and while helpful in its breadth of knowledge, I often found it overly "yes-man" like, and unwilling to engage me head on in any rigorous fashion. It just wasn't the same as discussing philosophy with a person.

    With that being said, one of the frustrations I've found with discussing philosophy, even with my close friends, is the sheer inability to remove oneself from one's own position. We are always responding in a manner that's a defense of our own position. When I try and talk to my Humean friend about Wittgenstein, despite respecting Witt, and being quite affluent in him, he rarely sees the point and thinks "Hume did it better" lol

    So, I wondered, what if A.I. could be both knowledgeable, and rigorous while not entrenched in any position - just there to reflect alongside you - not, as a replacement for engaging philosophically with others, but as a tool to help us reflect on our own positions more deeply, without getting bogged down in particular positions.

    At any rate, while working on the project, I started adding other things like music and animations, and after showing it to some friends, its grown into something that, I think, might actually help and be something other's would derive enjoyment from using.

    So, I'd like feedback on two things, really:

    1. What do you think of the philosophy, and direction of the project? Do you think A.I. has any "place" in philosophy?

    2. To any who would like to provide direct feedback to the project - let me know, and I'd be happy to PM a link to the website. The only semi private information it asks for is email, and that's just to mitigate abuse at this point. Every response costs me money, and while when you sign up, I let you use it for free, I can't allow limitless usage right now. I wont do anything with the email, and ill delete it from the system if you ever stop using the thing and want it removed. Currently I've only let my friends and those close to me use it, but I'd love to get more feedback and improve it. Idk, I've enjoyed working on it, and I use it myself and feel its helpful. It tries to visually map out your beliefs in a "space" you can zoom in and out and click on. There's a -currently buggy - avatar function and you can unlock items from interacting with it. Just little fun things that helped me stay engaged and felt fun.

    I hope this topic doesn't break any rules - but please let me know if it does and I'd be happy to remove/ modify it.
  • T Clark
    14.9k
    I started this project because, I tried using chat GPT to discuss philosophy, and while helpful in its breadth of knowledge, I often found it overly "yes-man" like, and unwilling to engage me head on in any rigorous fashion. It just wasn't the same as discussing philosophy with a person.013zen

    There are some metaphysical issues I've been thinking about. As part of that thinking, I made a list of metaphysical statements that seemed related and important to me, although I couldn't easily put into words how they were related. After thinking about them for a while, I punched them into Chat GPT and Gemini to see if that would give me anything to work with. The results I got were helpful. I went back and forth with the programs for a while to focus it's responses better. Since then, I've been using the results to focus my own understanding and come to my own conclusions.

    I plan to use the results of this work soon in a thread. I'll acknowledge AI's use, but the words and thoughts I use will be my own. I really don't have much of an opinion about AI use for more intensive and extensive philosophical work. I suggest you look at threads from @Pierre-Normand. He's put a lot of effort into this.

    With that being said, one of the frustrations I've found with discussing philosophy, even with my close friends, is the sheer inability to remove oneself from one's own position. We are always responding in a manner that's a defense of our own position.013zen

    But that's the whole point of discussing philosophy for me - to put your ideas out on the table for dissection. That's how you test them. That's how you find out how well you really understand your positions. It is a gift to have conscientious, intelligent friends who will try to rip your ideas apart. I didn't really use AI for that purpose in what I did. I used it more for synthesis than for analysis.

    So, I wondered, what if A.I. could be both knowledgeable, and rigorous while not entrenched in any position - just there to reflect alongside you - not, as a replacement for engaging philosophically with others, but as a tool to help us reflect on our own positions more deeply, without getting bogged down in particular positions.013zen

    At some level, that's what I did with my ideas, although I think you are probably talking about a more extensive effort than I used. As I see it, philosophy isn't about getting answers, it's about developing, expanding, and testing your own thoughts and understanding. Reading other philosophers, or using AI I suppose, can help with that, but it's work you have to do yourself.

    To any who would like to provide direct feedback to the project - let me know,013zen

    I'd be curious to take a look, but I'm not sure I'll give you the amount of feedback you seem to be looking for.
  • 013zen
    164
    With that being said, one of the frustrations I've found with discussing philosophy, even with my close friends, is the sheer inability to remove oneself from one's own position. We are always responding in a manner that's a defense of our own position.
    — 013zen

    But that's the whole point of discussing philosophy for me - to put your ideas out on the table for dissection. That's how you test them. That's how you find out how well you really understand your positions. It is a gift to have conscientious, intelligent friends who will try to rip your ideas apart. I didn't really use AI for that purpose in what I did. I used it more for synthesis than for analysis.
    T Clark

    I agree, and interestingly enough I also found myself ultimately using chatGPT exactly in a manner you described in your previous paragraph; more as an editor of my own ideas which I then refined through continual engagement - but, I never used it as an interlocutor again. I tried to make an interlocutor, because I agree with you - The heart of philosophy is putting forth your ideas and challenging them.

    But, what I'm describing is not this. My friends and I always have fruitful conversations, and I always learn a lot - mostly about their manners of approaching certain problems through the lense of their particular reading of so and so. While helpful in its own right, and where philosophy may "actually exist" in that it allows for the synthesis of different perspectives, I wondered if I could create an A.I that would challenge and engage - without having its own position, so to speak. Its goal being to reflect - openly - on whatever you want to reflect on.

    As I see it, philosophy isn't about getting answers, it's about developing, expanding, and testing your own thoughts and understanding. Reading other philosophers, or using AI I suppose, can help with that, but it's work you have to do yourself.T Clark

    I agree - and this project is meant to "try" and provide a space to do some of that work and still be engaging. It's designed to ask questions, more than it provides any answers or readings.....It may be a failure at that, but regardless of the project, I find I'll continue to use GPT to reflect on my own writing so see if, for example, I am being vague or if i am being too redundant or wordy, etc.

    I'd be curious to take a look, but I'm not sure I'll give you the amount of feedback you seem to be looking for.T Clark

    Anything is welcome and appreciated.
  • Count Timothy von Icarus
    3.9k


    I tried to make an interlocutor, because I agree with you - The heart of philosophy is putting forth your ideas and challenging them.

    It's too sycophantic is my problem with it. I really think that's one of the bigger risks of the technology. It just affirms whatever you feed to it, even praises it, and asking it to be critical doesn't seem to work that well. I imagine that partly this is done to keep customers happy.

    I have experimented with putting some manifestly stupid ideas to it and I still get fire emojis about my genius.
  • 013zen
    164
    It's too sycophantic is my problem with it. I really think that's one of the bigger risks of the technology. It just affirms whatever you feed to it, even praises it, and asking it to be critical doesn't seem to work that well. I imagine that partly this is done to keep customers happy.Count Timothy von Icarus



    Exactly what I mean when I describe GPT as "Yes-many" - this was my problem, too, and why I tried to make something better. Something that doesn't affirm, but questions and challenges. It still acknowledges your points, but isn't overly flattering or affirming. It wants to actively identify tensions in what you say, and discuss them.
  • Benkei
    8.1k
    This is every reason why it's a good editor and first draft writer. Other than that, it has no role in anything.
  • 013zen
    164


    As I've mentioned, this is something that I also found to be the case using AI like gpt... however, this, I think, mistakes accidental aspects of AI for necessary ones.

    Gpt is designed to be a yes man, but there's no reason to assume that an AI, by design, must lack rigor.

    This is why I created Pocket Philosopher - as a proof of concept. If you're interested, id be happy to pm the link and you can see if it's a step in the right direction.
  • Wayfarer
    24.6k
    1. What do you think of the philosophy, and direction of the project? Do you think A.I. has any "place" in philosophy?013zen

    absolutely. I've been interacting with ChatGPT and despite recognising that it is programmed to be positive about the user it's interacting with, it also comes with ideas and arguments that I hadn't considered. It's also very good at finding sources and citations for ideas.

    I've also signed up for (and paid for!) an app called Alter, after having heard a talk on it by John Vervaeke, who was involved in building it. I haven't had the chance to really assess it yet, but I think it's trying to do something similar to what you're up to.

    So, yes, very interested to see what you're working on.
  • 013zen
    164


    Interesting! Thanks for sharing the link, I'll definitely give this a look.

    Also, I PM you the link to my project. Any and all feedback is welcome. Thank you!!
  • Hanover
    13.9k
    It's too sycophantic is my problem with it.Count Timothy von Icarus

    I have found this as well, but I've also found that you can ask it not to be, and you can present your ideas as if they are being submitted by your opponent. You can even ask for a brutal attack of your positions, which can be pretty entertaining in its own right. Depending upon the version of AI you're using, it might be remembering your every post, and so you might want to ask it to brutally tell you what your limitations are, you can do that as well.

    In any event, we're just at the beginning of the AI revolution and I see the limitations we're pointing out as temporary, and they'll be altered over time in order to satisfy whatever the demand is. If what we actually want is brutal honesty, that will be coming.
  • Outlander
    2.4k
    It's too sycophantic is my problem with it.Count Timothy von Icarus

    You can actually make your own AI, that is to say, configure it entirely based on a few prompts or an entire library, with enough resource and where-with-all.

    There's a "fringe site" I visit on occasion (just for the news and on occasion to marvel at the depths of human depravity, or at least keep track of relevant trends associated with such) that did so and basically made it a "conspiracy AI" that pretty much suggests the opposite of what your standard AI will always. For what it's worth. Pretty sure that requires a fairly robust setup including your own web server and some sort of pricey license alongside considerably advanced technical knowledge. Or perhaps not. Not sure.

    Perhaps something the alleged, self-professed "web developer" @Jamal might wish to consider just as proof-of-concept one day? Imagine. "Feeding" it (AI) the entire written works of one philosopher (or several! perhaps conflicting one's!) and making it a free AI bot for us all to play around with. Now, that'd be going above and beyond, as a site owner. Bah. There has to be something like that already.

    (difference between regular AI is that when you "train" it based on only specific texts all it can do is rely on those texts whereas the standard AI relies on all available knowledge)
bold
italic
underline
strike
code
quote
ulist
image
url
mention
reveal
youtube
tweet
Add a Comment

Welcome to The Philosophy Forum!

Get involved in philosophical discussions about knowledge, truth, language, consciousness, science, politics, religion, logic and mathematics, art, history, and lots more. No ads, no clutter, and very little agreement — just fascinating conversations.

×
We use cookies and similar methods to recognize visitors and remember their preferences.