Loved it. I have this undeveloped idea that AI text generators, by virtue of being only trained on text, can at most generate what some philosophers would call "analytic knowledge", facts that are true only due to the relationship between words. Synthetic knowledge, in turn, requires accessing the experiential world. And that's where the real value comes from. I can share my experiences with you, and make you kinda experience it as well.
I'm not sure on the exact details myself, but we tend to project ourselves onto AI. Maybe we can call what it does analyzing or thinking, but we are not talking about the same process. When we conflate the two, we get into trouble.
Jun 19, 2023·edited Jun 19, 2023Liked by Lance Cummings
I have spent a small amount of time with two of these objects under discussion, Chat GPT, and Google's BARD. I used them knowing full well that the "responses" I get back to my queries are not natural or organically generated. With natural organic origin communications such as between two human beings first and foremost, from my perspective, to consider is the epistemological model of the two minds involved in any language interaction, the second consideration is understanding that all human expression of knowledge creation conveyed via some mode of communication ie. whether written, spoken, or signed (ASL etc.) take place within the larger context of the open system of reality, as opposed to the all too often erroneous assumption of communication taking place as always being confined to a closed system model, or empirically reducible.
With these two thoughts in mind I approach my games with the AI text generators with an intentional objective of eliciting something that is as close to a subjective response as is possible. I discovered that with one session I was able to essentially create an appearance of having gained a response that acknowledged the difference between an objective answer as opposed to a subjective answer.
It was interesting to observe that the written response I received once I made it clear to the program that it was permissible to provide a subjective answer the wording of the response was almost as if the AI text generator was pleased to be able to do so.
Interesting experiment! AI can seem to have intention but only mirrors our intention back at us. Intention isn’t in the words … it’s in the interaction.
Of course, we can say that of all text, but we need to be aware of the qualitative differences. AI has agency, but not the same agency as humans.
Indeed. The qualitative aspects are so easily overlooked by many in their desire to seek out and maybe meet something that has long been a staple of Science Fiction fans and those inspired by such imaginings to become the researchers, scientists, and other workers of technology.
I enjoy the way you seem to think as presented in your writings. While I do not have the same training or educational pedigree as you evidence, I will allow for myself the fact of my having been born with a significant hearing-loss that was undiscovered until I was seven-years old; the consequent result of which, after a lifetime of reflection and thought, I can say with the utmost confidence was to force me to become a genuinely focused listener and continuous observer of my environment, all of which has made me, if I may be so audacious as to say, a naturally developed savant when it comes to words and communications.
I appreciate technical terminologies that are developed and used in the many disciplines and specialized areas of study, unfortunately I have not and do not have either the time, patience, or inclination to subscribe to the typical means of knowledge acquisition and development that has been taken prisoner by the institutes of higher education milleaus.
I am just a another higher educational weed out, as Noam Chomsky describes- “The whole educational and professional training system is a very elaborate filter, which just weeds out people who are too independent, and who think for themselves, and who don't know how to be submissive, and so on -- because they're dysfunctional to the institutions.”
From my perspective and experience it is the higher educational systems that are somewhat dysfunctional; how else would the American Higher Educational System have produced such a crop of so-called highly educated individuals (Idiots in the most etymological sense of the word) who insist on supporting Trumpian type thinking, so much so that nearly half the voting population has and would again vote for such unthinking individuals, most of whom would not actually act in anyone's else's best interests but rather for their own profiting, controlling, biased interests?
Well said. For me, it's all about conveying ideas from one person to another (usually from my brain to another person's brain, or several other persons).
That's often called the transactional model of communication. I work from more of an interactional model. Ideas are built through our interaction between people, technology, and language (not from our inside our minds).
I'm curious ... how do you account for AI in this kind of transaction?
Interesting! I agree that ideas are built through interaction.
Maybe it's fair (and more accurate) to say that my personal synthesis and contribution to the discussion is best expressed through a pure idea, not through particular combinations of words. It's just that we're in an era where we have to combine alphanumeric characters to represent words, then combine words to make comprehensible phrases, and so on. I'm not convinced that era is going to last much longer, and I think that AI has a lot to do with it.
All of the synthesis, and what I want to contribute to the thoughts of another person, ultimately, reside in my head for a short period. I have to translate them into cohesive thoughts.
Loved it. I have this undeveloped idea that AI text generators, by virtue of being only trained on text, can at most generate what some philosophers would call "analytic knowledge", facts that are true only due to the relationship between words. Synthetic knowledge, in turn, requires accessing the experiential world. And that's where the real value comes from. I can share my experiences with you, and make you kinda experience it as well.
I'm not sure on the exact details myself, but we tend to project ourselves onto AI. Maybe we can call what it does analyzing or thinking, but we are not talking about the same process. When we conflate the two, we get into trouble.
I have spent a small amount of time with two of these objects under discussion, Chat GPT, and Google's BARD. I used them knowing full well that the "responses" I get back to my queries are not natural or organically generated. With natural organic origin communications such as between two human beings first and foremost, from my perspective, to consider is the epistemological model of the two minds involved in any language interaction, the second consideration is understanding that all human expression of knowledge creation conveyed via some mode of communication ie. whether written, spoken, or signed (ASL etc.) take place within the larger context of the open system of reality, as opposed to the all too often erroneous assumption of communication taking place as always being confined to a closed system model, or empirically reducible.
With these two thoughts in mind I approach my games with the AI text generators with an intentional objective of eliciting something that is as close to a subjective response as is possible. I discovered that with one session I was able to essentially create an appearance of having gained a response that acknowledged the difference between an objective answer as opposed to a subjective answer.
It was interesting to observe that the written response I received once I made it clear to the program that it was permissible to provide a subjective answer the wording of the response was almost as if the AI text generator was pleased to be able to do so.
Interesting experiment! AI can seem to have intention but only mirrors our intention back at us. Intention isn’t in the words … it’s in the interaction.
Of course, we can say that of all text, but we need to be aware of the qualitative differences. AI has agency, but not the same agency as humans.
Indeed. The qualitative aspects are so easily overlooked by many in their desire to seek out and maybe meet something that has long been a staple of Science Fiction fans and those inspired by such imaginings to become the researchers, scientists, and other workers of technology.
I enjoy the way you seem to think as presented in your writings. While I do not have the same training or educational pedigree as you evidence, I will allow for myself the fact of my having been born with a significant hearing-loss that was undiscovered until I was seven-years old; the consequent result of which, after a lifetime of reflection and thought, I can say with the utmost confidence was to force me to become a genuinely focused listener and continuous observer of my environment, all of which has made me, if I may be so audacious as to say, a naturally developed savant when it comes to words and communications.
I appreciate technical terminologies that are developed and used in the many disciplines and specialized areas of study, unfortunately I have not and do not have either the time, patience, or inclination to subscribe to the typical means of knowledge acquisition and development that has been taken prisoner by the institutes of higher education milleaus.
I am just a another higher educational weed out, as Noam Chomsky describes- “The whole educational and professional training system is a very elaborate filter, which just weeds out people who are too independent, and who think for themselves, and who don't know how to be submissive, and so on -- because they're dysfunctional to the institutions.”
From my perspective and experience it is the higher educational systems that are somewhat dysfunctional; how else would the American Higher Educational System have produced such a crop of so-called highly educated individuals (Idiots in the most etymological sense of the word) who insist on supporting Trumpian type thinking, so much so that nearly half the voting population has and would again vote for such unthinking individuals, most of whom would not actually act in anyone's else's best interests but rather for their own profiting, controlling, biased interests?
A great article that elucidates the real meaning behind the art of writing.
Well said. For me, it's all about conveying ideas from one person to another (usually from my brain to another person's brain, or several other persons).
That's often called the transactional model of communication. I work from more of an interactional model. Ideas are built through our interaction between people, technology, and language (not from our inside our minds).
I'm curious ... how do you account for AI in this kind of transaction?
Interesting! I agree that ideas are built through interaction.
Maybe it's fair (and more accurate) to say that my personal synthesis and contribution to the discussion is best expressed through a pure idea, not through particular combinations of words. It's just that we're in an era where we have to combine alphanumeric characters to represent words, then combine words to make comprehensible phrases, and so on. I'm not convinced that era is going to last much longer, and I think that AI has a lot to do with it.
All of the synthesis, and what I want to contribute to the thoughts of another person, ultimately, reside in my head for a short period. I have to translate them into cohesive thoughts.