ChatGPT answers metaphysical questions :)

[AshvinP]

The tool can not only respond with written text but can also generate a short audio, podcast-style summary of the sources. I must admit I was quite blown away by this feature. Not only that the content itself makes sense but also the quality of the generated voices is impressive. To be fair, if I had heard that without knowing, I wouldn’t have guessed that it was generated.

Here’s an example: Link 1

Here are two more versions generated over earlier versions of the text: Link 2, Link 3

The voice quality and the articulation are really almost scary to behold.

Indeed, the resemblance with the auditory expression of human reasoning and dialogue is impressive.

I guess this is spiritually very helpful, in that it puts under our nose the demonstration that the linguistic (word-conceptual) layer in which we operate verbally (in thought and speech) has mostly become in our times so abstract and disconnected from meaningful gesture and meaningful sound, that we can hardly make the difference between an expression of truly human spiritual activity and a statistically significant but fake patching of verbal sequences, be it in speech or in text form.

Our thinking voice and physical voice (our language) rarely convey our intuitive human potential. It is rather the expression of our dwelling in a sort of internally coherent, but parallel and disconnected layer of ideal existence, and can therefore be easily hacked. If we were to reconnect language with feeling and to reinfuse it with music and gesture, the LLMs would read and sound like childish caricatures of an authentically human voice, because we would be able to both express ourselves in a much more pregnant language, and to detect that character in our fellow humans.

Yeah, I came across one of these podcasts on another philosophical-spiritual Substack post and, even though it made little sense that a real podcast would have been made about this little-known post, I totally thought it was two real people Of course, now that I know this technology is out there, it will be easier to differentiate, especially if the same male/female voices are used. That makes me wonder how exactly that works - from whence are the voices sourced and how are they generated? Is it some unique voice constructed from a sampling of many other podcasts? To what extent can well-known voices be replicated speaking content they have never actually spoken before?

As said, it is somewhat relieving that if nothing else, the words are at least mechanically fitting together. I could refer to this next time when there’s an argument about the random noise that inner phenomenology consists of. I hope that we can then move the conversation in a region where it should really belong. It’s like saying: “Look, blaming that the words don’t even make mechanical sense, is not really plausible. This can be seen from the fact that a computational algorithm, which doesn’t suffer from laziness or irrational antipathy toward the text, can pretty well match the numerical patterns against the statistical database of coherent grammatically sound human language. With this out of the way, let’s focus on the real issues – these inner forces that prevent us from approaching the meaning and its living experience, by trying to convince us that there’s no reason to even try read the text because we have apriori assumed that it makes not even grammatical sense.”

This is a very creative way to address this particular objection! Even if it doesn't have much impact, it is at least heartening to see how these relatively small yet redemptive functions of AI tech can gradually start to emerge through our creative thinking.

[Cleric]

Our thinking voice and physical voice (our language) rarely convey our intuitive human potential. It is rather the expression of our dwelling in a sort of internally coherent, but parallel and disconnected layer of ideal existence, and can therefore be easily hacked. If we were to reconnect language with feeling and to reinfuse it with music and gesture, the LLMs would read and sound like childish caricatures of an authentically human voice, because we would be able to both express ourselves in a much more pregnant language, and to detect that character in our fellow humans.

Very true. Our compressing elemental nature in a sense is already something like a statistically weighed database, even though of a much different kind than the flat computational model. As such, it really is being activated by the most various 'winds', and semi-automatic verbal sequences are spat out. The scary thing is that our present educational system actively cultivates such elemental (stored) intelligence.

Yeah, I came across one of these podcasts on another philosophical-spiritual Substack post and, even though it made little sense that a real podcast would have been made about this little-known post, I totally thought it was two real people Of course, now that I know this technology is out there, it will be easier to differentiate, especially if the same male/female voices are used. That makes me wonder how exactly that works - from whence are the voices sourced and how are they generated? Is it some unique voice constructed from a sampling of many other podcasts? To what extent can well-known voices be replicated speaking content they have never actually spoken before?

I don't know the technical details (I don't think ggl have disclosed such information yet) but there are many things out there that give us hints. Replicating voices has been mastered already pretty well. They need only a 3-second sample of your voice to extract your timbre fingerprint and then generate any speech (for example, in text-to-speech manner).

There are now even near real-time voice transformers (example).

About the podcasts - I guess they have trained the style and the specific articulations on tons of real podcasts. The voices themselves are a matter of choice probably. The summaries are easily generated with the existing Gemini model as text. One possibility is that they simply pass that text through text-to-speech engine, but I'm not sure at what stage the specific articulations are added and from where.

[Cleric]

Here's a nice short overview of language models. Of course, the more technical details are vastly abstracted away, but at least in common lines, it can give orientation, especially for people new to the tech.

[AshvinP]

Here's a nice short overview of language models. Of course, the more technical details are vastly abstracted away, but at least in common lines, it can give orientation, especially for people new to the tech.

Thanks, this intro is helpful. I was just watching an interview with JP where he mentions LLMs and describes them as providing a 'hard science' of linguistic cognition and perception, particularly its inherently symbolic function for anchoring 'micro-narratives'. In that sense, the LLMs can be very useful mirrors for how we are normally interacting with our symbolic perceptual environment, as long as we don't take in a strictly mechanical-computational way.

[Federica]

Here's a nice short overview of language models. Of course, the more technical details are vastly abstracted away, but at least in common lines, it can give orientation, especially for people new to the tech.

Thinking about Levin's view on emergence, it would be interesting to ask the author what he means by "emergent phenomenon":

"Although researchers design the framework for how each of these steps work, it's important to understand that the specific behavior is an emergent phenomenon based on how those hundreds of billions of parameters are tuned during training.
This makes it incredibly challenging to determine why the model makes the exact predictions that it does.
What you can see is that when you use large language model predictions to autocomplete a prompt, the words that it generates are uncannily fluent, fascinating, and even useful."

[Federica]

Thanks, this intro is helpful. I was just watching an interview with JP where he mentions LLMs and describes them as providing a 'hard science' of linguistic cognition and perception, particularly its inherently symbolic function for anchoring 'micro-narratives'. In that sense, the LLMs can be very useful mirrors for how we are normally interacting with our symbolic perceptual environment, as long as we don't take in a strictly mechanical-computational way.

It seems to me JP mixes up the "symbolic world" with the world of ideas. He doesn't distinguish the word-symbol from the idea, as if they were a direct reflection of one another. It doesn't occur to him that, to paraphrase Steiner, the vast majority of people nowadays don't really think, in full possession of intuitive gestures. We think in words, we let abstractified words, spiritually divorced words, drag our trains of thought and shape their form from without, semi-automatically. And so the linguistic patterns the LLM train on are 99,9% those calcified sequences of abstractified symbols. JP flattens two planes on one plane. He literally confuses the concept with the word, at 26:56.

Therefore, his understanding is that "LLM perfectly model cognition". But they don't. They do illustrate and reveal, in their makeup, the quality of certain human cognitive patterns, but I don't think they model cognition. They model the outer expression of our symbol-addicted, life-less, and chain-like linguistic habits. Chain-like to mean both like in a prison, and like in a concatenation. (This is not a reproach to JP, of course I can only notice that thanks to exposure to phenomenology and the path of thinking).

On the other hand, his definition of perceptions as micro-narratives is really good, in a sense. Though one could take it by its drawback, that a narrative is something that lives inside our mind. It seems to me there's a lot of boiling material in what he says next, but the connections are not stable. There would be much to say about all this firing of ideas, in any case, I see again the concept of pyramid, which was there also in the other dialog you recently posted with him. He speaks structure, but he thinks pyramid and draws one with his arms Also there seems to be a few drops of materialistic thinking in the mix, when he speaks of those structures in the collective unconscious (suprime stories) in connection with reproduction/biology. tbc...

[AshvinP]

It seems to me JP mixes up the "symbolic world" with the world of ideas. He doesn't distinguish the word-symbol from the idea, as if they were a direct reflection of one another. It doesn't occur to him that, to paraphrase Steiner, the vast majority of people nowadays don't really think, in full possession of intuitive gestures. We think in words, we let abstractified words, spiritually divorced words, drag our trains of thought and shape their form from without, semi-automatically. And so the linguistic patterns the LLM train on are 99,9% those calcified sequences of abstractified symbols. JP flattens two planes on one plane. He literally confuses the concept with the word, at 26:56.

Therefore, his understanding is that "LLM perfectly model cognition". But they don't. They do illustrate and reveal, in their makeup, the quality of certain human cognitive patterns, but I don't think they model cognition. They model the outer expression of our symbol-addicted, life-less, and chain-like linguistic habits. Chain-like to mean both like in a prison, and like in a concatenation. (This is not a reproach to JP, of course I can only notice that thanks to exposure to phenomenology and the path of thinking).

On the other hand, his definition of perceptions as micro-narratives is really good, in a sense. Though one could take it by its drawback, that a narrative is something that lives inside our mind. It seems to me there's a lot of boiling material in what he says next, but the connections are not stable. There would be much to say about all this firing of ideas, in any case, I see again the concept of pyramid, which was there also in the other dialog you recently posted with him. He speaks structure, but he thinks pyramid and draws one with his arms Also there seems to be a few drops of materialistic thinking in the mix, when he speaks of those structures in the collective unconscious (suprime stories) in connection with reproduction/biology. tbc...

The thing to keep in mind with JP is he is always speaking phenomenologically, using concepts to trace certain undeniable experiential relations. He isn't doing any sort of metaphysics or trying to exhaustively explain the "nature of cognition" with the LLM example. For him, as it also is for us, LLM is only an analogical symbol for what he has noticed of some of the characteristic movements by which ordinary cognition interfaces with its perceptual environment, which of course must be the case since LLM is developed and trained by ordinary cognition and precisely for the aim of making it more closely imitate those cognitive movements. It is exactly what you say in bold, that's all. I think Cleric's video reveals some of the details of how that is the case as well. It is very much what JP illustrates with the example of 'witch' evoking a certain 'cloud of images and words' in a 'statistically associated' way. This is simply an artistic way of describing some aspect of how we experience it.

This phenomenological thinking becomes more evident the more we work through the expanded context of his ideas, as expressed through many recent discussions. Even later in this same video it becomes clear that he understands the Spirit that animates our cognitive movements as mysterious and irreducible (cannot be categorized), so he is not flattening all of that to the LLM intellectual plane. When he speaks about how our imaginative (or 'abstract') thinking activity has come to feedback into reproduction/biology by allowing us to play out ideas in the theatre of imagination before implementing them, again, he is pointing to undeniable experiential facts of our individual and collective existence without speculating on any metaphysical considerations of 'what reproduction/biology actually is' or 'what cells are made of' and so forth. It also comes out in the discussion of what it means to "believe in God" and his critique of the 'protestant propositional approach' in which mere sequences of words (doctrines and creeds) have become confused for the inner realities they should be prompting us to engage with more intimately.

In his thinking, the 'statistical weighting of ideas' which inform our meaningful narratives always stems from our ideals and values, which again he feels can be supported empirically, and he often links that kind of ideal-oriented spiritual activity to the primal structuring of order from chaos (spoken of in ancient myth like Genesis). All of this can be very fruitful for our orientation to spiritual science if we simply treat it as a phenomenology of ordinary imaginative life and try to sense the underlying inner gestures, the core intuitions. Becoming more sensitive and conscious to these connections, as JP is doing through his thoughtful explorations, is exactly what brings life and higher meaning into our chain-like thinking habits. We are revealing more of the symbolic-imaginative structure that is always implicit in those thinking-perceptual movements but is normally too dulled out by our narrow focus on the content of our thoughts and perceptions, and that is exactly what loosens the chains and redeems thinking from the inside-out.

PS - I really dont have interest in arguing about who is right or wrong about what JP specifically thinks in terms of reductionism and so forth. The non-reductive thinking is very evident to me from my in depth exploration of his ideas, especialy in recent months when it is fresh in my mind. I only offer this response to stimulate thinking about the core intuitions expressed, and the differences between depth and planar thinking. If particular words or phrases used in the video start getting dissected, count me out

[Federica]

The thing to keep in mind with JP is he is always speaking phenomenologically, using concepts to trace certain undeniable experiential relations. He isn't doing any sort of metaphysics or trying to exhaustively explain the "nature of cognition" with the LLM example. For him, as it also is for us, LLM is only an analogical symbol for what he has noticed of some of the characteristic movements by which ordinary cognition interfaces with its perceptual environment, which of course must be the case since LLM is developed and trained by ordinary cognition and precisely for the aim of making it more closely imitate those cognitive movements. It is exactly what you say in bold, that's all. I think Cleric's video reveals some of the details of how that is the case as well. It is very much what JP illustrates with the example of 'witch' evoking a certain 'cloud of images and words' in a 'statistically associated' way. This is simply an artistic way of describing some aspect of how we experience it.

This phenomenological thinking becomes more evident the more we work through the expanded context of his ideas, as expressed through many recent discussions. Even later in this same video it becomes clear that he understands the Spirit that animates our cognitive movements as mysterious and irreducible (cannot be categorized), so he is not flattening all of that to the LLM intellectual plane. When he speaks about how our imaginative (or 'abstract') thinking activity has come to feedback into reproduction/biology by allowing us to play out ideas in the theatre of imagination before implementing them, again, he is pointing to undeniable experiential facts of our individual and collective existence without speculating on any metaphysical considerations of 'what reproduction/biology actually is' or 'what cells are made of' and so forth. It also comes out in the discussion of what it means to "believe in God" and his critique of the 'protestant propositional approach' in which mere sequences of words (doctrines and creeds) have become confused for the inner realities they should be prompting us to engage with more intimately.

In his thinking, the 'statistical weighting of ideas' which inform our meaningful narratives always stems from our ideals and values, which again he feels can be supported empirically, and he often links that kind of ideal-oriented spiritual activity to the primal structuring of order from chaos (spoken of in ancient myth like Genesis). All of this can be very fruitful for our orientation to spiritual science if we simply treat it as a phenomenology of ordinary imaginative life and try to sense the underlying inner gestures, the core intuitions. Becoming more sensitive and conscious to these connections, as JP is doing through his thoughtful explorations, is exactly what brings life and higher meaning into our chain-like thinking habits. We are revealing more of the symbolic-imaginative structure that is always implicit in those thinking-perceptual movements but is normally too dulled out by our narrow focus on the content of our thoughts and perceptions, and that is exactly what loosens the chains and redeems thinking from the inside-out.

PS - I really dont have interest in arguing about who is right or wrong about what JP specifically thinks in terms of reductionism and so forth. The non-reductive thinking is very evident to me from my in depth exploration of his ideas, especialy in recent months when it is fresh in my mind. I only offer this response to stimulate thinking about the core intuitions expressed, and the differences between depth and planar thinking. If particular words or phrases used in the video start getting dissected, count me out

My goal is to understand language in depth, therefore I am not attached to my impressions about JPs trains of thought. Though the bold is not entirely clear to me, for example in connection to how he references the Collective Unconscious in this clip, I am not very familiar with JP and I don't need to find out now. I am sure you have a better understanding of his mind. Unrelated to that, another question coming from reading your post is: what is the deeper meaning of giving warnings? But I also don't need to find out now. As I said, my focus is on language, and, contingently, on LLMs as a concrete opportunity or angle along which one can enter the question.

[Cleric]

Thinking about Levin's view on emergence, it would be interesting to ask the author what he means by "emergent phenomenon":

"Although researchers design the framework for how each of these steps work, it's important to understand that the specific behavior is an emergent phenomenon based on how those hundreds of billions of parameters are tuned during training.
This makes it incredibly challenging to determine why the model makes the exact predictions that it does.
What you can see is that when you use large language model predictions to autocomplete a prompt, the words that it generates are uncannily fluent, fascinating, and even useful."

Knowing Grant Sanderson from interviews (he has two podcasts with Lex, for example), I would say he is a very down-to-earth thinker, exceptionally clear and concise, with an enviable pedagogical approach, always striving to build intuitive understanding, yet staying firmly within the sensory-intellectual field. From some questions that Lex asked him, it was clear that he has no interest in speculating beyond mathematics and physics (like in this and this chapter). With that in mind, I'm pretty sure he means nothing mystical about the emergent behavior. I think he would agree that in this case it is a synonym of 'surprising' or 'unexpected', something which we did not code specifically for, but nevertheless proceeds from the statistics.

[AshvinP]

My goal is to understand language in depth, therefore I am not attached to my impressions about JPs trains of thought. Though the bold is not entirely clear to me, for example in connection to how he references the Collective Unconscious in this clip, I am not very familiar with JP and I don't need to find out now. I am sure you have a better understanding of his mind. Unrelated to that, another question coming from reading your post is: what is the deeper meaning of giving warnings? But I also don't need to find out now. As I said, my focus is on language, and, contingently, on LLMs as a concrete opportunity or angle along which one can enter the question.

It may be useful to ask in these situations where the meaning is ambiguous, if we were using 'collective unconscious' (for example) as a symbol for phenomenological realities, how would we understand its meaning? I know that I have often come across this same situation where I am unsure if a thinker means something in one way or another, and it's easy to assume they are flowing along with abstract thinking habits (usually they are). Lately, unless a person is explicitly weaving together metaphysical theories, I try to give the benefit of the doubt and seek the ways in which their concepts can be understood as phenomenological descriptions. If more people gave that same benefit when approaching spiritual science in the archive or on this forum, a lot of misunderstandings and skepticism would be cleared up.

As for JP, it occurred to me this elegantly expressed passage from Part 6 basically sums up his entire inner stance. In fact, I think that if he came across this passage without any of the previous context, he would still basically understand its meaning and how it overlaps with his own deep intuitions. He has even used the musically nested spheres of activity (personal, family, community, nation, etc.) analogy before.

It is inevitable that the incoming flow can initially be grasped only through that which we are already familiar with. In a sense, our compressed kernel serves as the ‘lowest common denominator’. It is for this reason that when we think of the more encompassing forms of inner activity, which can bend the spacetime flow at greater scales, initially we can only grasp them as far as they intersect with the images of our meso-scale bodily life. For example, when we awaken to our inner being, we realize that we have the inner freedom to support a lifelong curvature of our existential flow. This is what usually has been achieved in the ages by having a certain religious orientation in life. In today’s superficial age, we dimly steer through the consumption of condensing mental images, feelings, and sensations. There’s very little in our materialistic culture that can give us a greater-scale intuitive orientation within our flow. The genuine religions of the past have aimed to instill such curvatures into human destinies by inspiring them through scaled images of Divine spacetime curvatures of the World flow. Today we can approach this in a much more conscious way. Such a lifelong curvature can serve as our firm moral determination to always grow in consciousness of what is good and beautiful, and allow it to flow out through our will, thus striving for an individual flow that musically nests within the total World flow and as such, benefitting and ennobling it. This is not something that we set in place once and then leave on autopilot. We need constant vigilance to support that ideal curvature in the ‘now’. It is as if we have intended to count but this intention is spread over our whole life span and even beyond, and at any point of our daily life, no matter what kind of momentary Tetris pieces condense at our perceptual horizon, we can remember “I’m counting” – that is: “I don’t forget that I strive to always guide my flow in harmony with my freely intended lifelong curvature, such that at any point I can give a meaningful trajectory of my existential movie. In this way, even the smallest thoughts and acts can be experienced as meaningfully nested within a great musical contextuality.” This lifelong moral curvature is not something fixed but a high ideal striving for the ever-evolving intuition of what we are, how we are embedded in the World flow, what that World flow is doing on all scales, how our activity reverberates in it, and how we can creatively contribute to the total symphony. Thus, at the beginning this moral curvature that we support at the great spacetime scales can only be comprehended as far as it finds its reflection in the meso-scale images of bodily life – what we do, what we say, and so on. But as our intuition expands we gradually understand that concealed in these flow-bending moral forces are possibilities that in time can help humanity as a whole to artistically shape and steer the collectively harmonized spacetime flow in ways that presently are in the most literal sense unimaginable. This collective harmonization is not a formal agreement to follow the decrees of some utopic ideological system, but the result of our innermost beings finding each other within the Solar Cosmic Spheres of inner space, from whence the voice of conscience speaks.

Also, I'm not sure what you were referring to with 'giving warnings'?