bbenzon

I gave ChatGPT three lists of topics for which it had to propose categories into which the topics could be sorted. Two lists were relatively short, 56 and 53 topics; I asked ChatGPT to propose six organizing categories for each. One list was much longer, 655 topics; I asked ChatGPT to propose 12 for categories for it. In all cases the proposed categories were reasonable. ChatGPT explained each proposed category either with a pair of sentences (the short lists), or with characterizing phrases (the long list). These characterizations were reasonable. In a further task, when asked to place topics under the proposed categories, ChatGPT placed many topics under the first two categories and very few under the last two. Though quite different in detail, this task has a rough formal similarity to generating a coherent story that is organized on three levels: 1) the whole story, 2) story segments, 3) sentences in story segments.

ChatGPT responds the prompt, "story", with a simple story. 10 stories elicited by that prompt in a single session have a greater variety of protagonists than 10 stories each elicited in its own session. Prototype: 19 stories were about protagonists who venture into the world and learn things that benefit their community. ChatGPT's response to that simple prompt gives us a clue about the structure of the underlying model.

In a few cases ChatGPT responds to a prompt (e.g. "To be or not to be") by returning a specific text word-for-word. More often (e.g. "Johnstown flood, 1889") it returns with information, but the specific wording will vary from one occasion to the next. In some cases (e.g. "Miriam Yevick") it doesn't return anything, though the topic was (most likely) in the training corpus. When the prompt is the beginning of a line or a sentence in a famous text, ChatGPT always identifies the text. When the prompt is a phrase that is syntactically coherent, ChatGPT generally identifies the text, but may not properly locate the phrase within the text. When the prompt cuts across syntactic boundaries, ChatGPT almost never identifies the text. But when told it is from a "well-known speech" it is able to do so. ChatGPT's response to these prompts is similar to associative memory in humans, possibly on a holographic model.

This article recounts an intellectual journey that began in curiosity about the structure of Coleridge's "Kubla Khan" in the late 1960s and has led to an interest in large language models at the present time. A close analysis of the poem revealed its two parts each to have a nested structure (think of a matryoshka doll) that suggested the operation of an underlying computational process (nested loops). That led to the study of computational linguistics (semantic networks), followed by neuroscience (Karl Pribram's neural holography), and cultural evolution. In the 2010s I began following work digital humans had been doing with machine learning. When GPT-3 was released in 2020 I was ready, though it took me awhile to establish a link, however tentative, between that conceptual universe and that of "Kubla Khan."

Culture elaborates expressive forms by developing ever more differentiated control over patterns in the expressive medium. In Rank 1 culture (preliterate) music evolves through control over rhythm. Rank 2 culture (literacy) gains control over melodic structure while Rank 3 culture (Renaissance and after) adds harmonic elaboration to rhythm and melody. Within the twentieth century jazz has followed a similar course, with rhythmic elaboration coming first with traditional jazz, then melodic control emerging with swing, and harmonic control with bop. Both classical and jazz have much music straining beyond the limits of Rank 3 harmonic control, but no clear Rank 4 forms have yet evolved.

Chuck Jones believed in the “disciplines” one had to maintain for a cartoon, the constraints within which one acted. In the Road Runner cartoons, no one talked, though there could be signage, the action always centered on two, and only two, characters, Road Runner and Wile E. Coyote, and it always takes place in the outdoors in the American Southwest. These cartoons play with reality and engage in a guessing game with the audience in doing so. What’s Opera, Doc? goes beyond a string of gags end engages in a story about the relationship between Elmer Fudd and Bugs Bunny. Finally Daffy Duck and Porky Pig venture into space in Duck Dodgers in the 24½th Century, which satirizes incompetence and self-generating conflict.

This is a brief guide to the various papers and books that William Benzon and David Hays have written about the long-term evolution of culture. Much of the work is descriptive in that it characterizes a variety of cultural phenomena at four different cultural Ranks as characterized by the conceptual mechanisms available to the culture but does not attempt to describe the causal process by which these mechanism have evolved in human history. The ranks have numerical designations that align with widely recognized historical epochs: Rank 1, preliterate; Rank 2, literacy; Rank 3, post-Renaissance West, industrialism; and Rank 4, 20th century. Ranges of phenomena considered: basic cognition and thinking, narrative, music, expressive culture, technology, forms of governance and economic organization. The approach is briefly contrasted with work by Robert Wright, Richard Dawkins, Boyd and Richerson, and others.

This is a dialog between a Naturalist Literary Critic and a Sympathetic Techno-Wizard about the interaction of symbols and neural nets in understanding "Kubla Khan," which has an extraordinary structure. Each of two parts is like a matryoshka doll nested three deep, with the last line of the first part being repeated in the middle of the second. They start talking about traditional symbol processing, with addressable memory, and nested loops, and end up talking about a pair of interlinked neural nets where one (language forms) is used to index the other (meaning).

These notes are in preparation for a post I will make on 5 July 2010 at the Forum on the website the National Humanities Center. I have been posting them on my blog, New Savanna (see URL below). My Forum post will, of course, be an independent and free-standing document; but it will be based largely on ideas in these posts and in the three further posts I am planning to do in the “workshop” phase of this project. It is possible, of course, that my ideas will change when writing the Forum post, but I do not anticipate drastic changes. Note: I have added a new post on memes to version 2: RESET: Where Are Memes? It modifies the position I elaborated in CE 3, 4, 5, and 8.

Opening paragraph: The ontology of common sense is the discipline which seeks to establish the categories which are used in everyday life to characterize objects and events. In everyday life steel bars and window panes are solid objects. For the scientist, the glass of the window pane is a liquid, and the solidity of both the window pane and the steel bar is illusory, since the space they occupy consists mostly of empty regions between the sub-atomic particles which constitute these objects. These facts, however, have no bearing on the ontological categories of common sense. Sub-atomic particles and solid liquids do not exist in the domain of common sense. Common sense employs different ontological categories from those used in the various specialized disciplines of science.

Cultural Studies, Future Studies, Artificial Intelligence, Information Technology, Futurism ...more ▾ With cultural evolution new processes of thought appear. Abstraction is universal, but rationalization first appeared in ancient Greece, theorization in Renaissance Italy, and model building in twentieth-century Europe. These processes employ the methods of metaphor, metalingual definition, algorithm, and control, respectively. The intellectual and practical achievements of populations guided by the several processes and exploiting the different mechanisms differ so greatly as to warrant separation into cultural ranks. The fourth rank is not completely formed, while regions of the world and parts of every population continue to operate by the processes of earlier ranks. It is the continued emergence of fourth rank and even fifth rank cognition that constitutes the coming singularity, not the emergence of “superintelligent” computers, whatever they may be.

This document collects a series of blog posts in which I prompted ChatGPT to tell a story. The exact nature of the prompts varied from simple-"Tell me a story"-to complex, where ChatGPT had to create a new story based on an existing story given in the prompt. Many of the stories were like fairy tales, some were realistic, and some where even about real people. When asked to tell a story about colorless green ideas, ChatGPT refused, and it refused to include an incident about a flying saucer in and ongoing fairy-tale kind of story. It told stories about many crimes, but not all crimes (no murders or rapes). I list 12 occasions when ChatGPT's response surprised me. The posts were uploaded between January 16 and February 21, 2023.

People in oral cultures tell stories as a source of mutual knowledge in the game theory sense (think: “The Emperor Has No Clothes”) on matters they cannot talk about either because they resist explicit expository formulation or because they are embarrassing and anxiety provoking. The communal story is thus a source of shared value and mutual affirmation. And the academic profession of literary criticism came to see itself as a repository of that shared value. Accordingly, in the middle of the 20th century it turned toward interpretation as its central activity. But critics could not agree on interpretations and that precipitated a crisis that led to Theory. The crisis has quited down, but is not resolved.

Kawajiri’s Ninja Scroll features two interlaced of sex scenes, each involving men who are large and hulking and men who are smaller. But one man is clothed in the robes of minor officialdom and has appropriately dressed hair while the other is semi-naked and bald. The two scenes, while both in dim night-time light, are depicted in different colors and the scenes are composed in contrasting ways. One of these acts is legitimate (the one involving the official) in the sense that it accords with social standards. The other is clearly rape. The film reveals both to be impure and the social order to be corrupt.

Shakespeare’s Sonnet 129 is considered in two analytical contexts. The main context is as a case study in abstract cognition as studied by David Hays and his students. In that model a three level cognitive system (systemic, episodic and gnomonic) is linked to the world through a sensorimotor system (thus making the model one of embodied cognition). As an example, the concept of lying (“…lust/ Is perjur’d…”) is constructed within a network that represents the relationship between an utterance and some state of affairs in the world (figures 4 and 5). The essay then introduces another context, that of a cognitive system embedded in a brain whose activity is realized though complex electrochemical processes. The situation depicted in the poem is then described as that arising in a neurally embodied cognitive network where one set of cognitive capacities is active in anticipation (one neurochemical state) and pursuit of sexual satisfaction while a different, and incompatible, set of capacities (a different neural state) is active after satisfaction has faded into the past. The poem itself is a contemplative record of that maddening cyclic action.

After an introductory chapter, Mithen has six chapters that are more or less about music and the brain. The second part of the book consists of ten chapters starting with a review of primate communication and then presenting Mithen’s speculative reconstruction of the steps leading through the origins and elaboration of musicking to its final differentiation into language and music. All of these chapters contain useful information and intelligent synthesis that is well and cleanly presented to a cumulatively brilliant effect. Rather than attempt to review all of this material I will only cover the topics that, for better or worse, I found interesting. I will then offer some speculations of my own on music and sociality and conclude with some general remarks.

In The Wind Rises Hayao Miyazaki weaves various modes of experience in depicting the somewhat fictionalized life of Jiro Horikoshi, a Japanese aeronautical engineer who designed fighter planes for World War II. Horikoshi finds his vocation through ‘dreamtime’ encounters with Gianni Caproni and courts his wife with paper airplanes. The film opposes the wind and chance with mechanism and design. Horikoshi’s attachment to his wife, on the one hand, and to his vocation on the other, both bind him to Japan while at the same time allowing him to separate himself, at least mentally, from the imperial state.

While science has accepted biological evolution through natural selection, there is no generally agreed explanation for why evolution leads to ever more complex organisms. Evolution yields organismic complexity because the universe is, in its very fabric, inherently complex, as suggested by Ilya Prigogine's work on dissipative structures. Because the universe is complex, increments in organismic complexity yield survival benefits: (1) more efficient extraction of energy and matter, (2) more flexible response to vicissitudes, (3) more effective search. J.J. Gibson's ecological psychology provides a clue to the advantages of sophisticated information processing while the lore of computational theory suggests that a complex computer is needed efficiently to perform complex computations (i.e. sophisticated information processing).

Osamu Tezuka’s Metropolis is ultimately a narrative about the nature of art told in a concrete visual and verbal logic accessible to a child. The story unfolds across multiple domains – Greater Metropolis, Underground, Child’s World – and has the form of a ring where the first element matches the last, the second matches the next to last, and so forth: 1 2 3 4 5 4’ 3’ 2’ 1’. By juxtaposing multiple plot lines, Tezuka deploys an essentialist mechanics of inner-workings and external appearances of people and animals to illustrate practical, moral, and aesthetic issues. By depicting a world that is rapidly destroying itself, Tezuka indicates the order that must be conserved if the world is to persist. Two of Tezuka’s devices have been developed in later manga by himself and others: 1) an artificial being is search of its place in human society, 2) a being that can be either male or female.

In his work on memetics Daniel Dennett does a poor job of negotiating the territory between philosophy and science. The analytic tools he has as a philosopher aren’t of much use in building accounts of the psychological and social mechanisms that underlie cultural processes. The only tool Dennett seems to have at his disposal is analogy. That’s how he builds his memetics, by analogy from biology on the one hand and computer science on the other. These analogies do not work very well. To formulate an evolutionary account of culture one needs to construct one’s gene and phenotype analogues directly from the appropriate materials, neurons and brains in social interaction. Dennett doesn’t do that. Instead of social interaction he has an analogy to apps loading into computers. Instead of neurons he has homuncular agents that are suspiciously like his other favorite homuncular agents, memes. It doesn’t work.

Russell A. Barkley has argued that ADHD is fundamentally a disorientation in time. These notes explore the possibility that music, which requires and supports finely tuned temporal cognition, might play a role in ameliorating ADHD. The discussion ranges across cultural issues (grasshopper vs. ant, lower rate of diagnosis of ADHD among African-Americans), play, distribution of dopamine and norepinephrine in the brain, neural development, and genes in culture (studies of the distribution of alleles for dopamine receptors). Unfortunately, the literature on ADHD does not allow us to draw strong conclusions. We do not understand what causes ADHD nor do we understand how best to treat the condition. However, in view of the fact that ADHD does involve problems with temporal cognition, and that music does train one’s sense of timing, the use of music therapy as a way of ameliorating ADHD should be investigated. I also advocate conducting epidemiological studies about the relationship between dancing and music in childhood, especially in early childhood, and the incidence of ADHD.

While science has accepted biological evolution through natural selection, there is no generally agreed explanation for why evolution leads to ever more complex organisms. Evolution yields organismic complexity because the universe is, in its very fabric, inherently complex, as suggested by Ilya Prigogine's work on dissipative structures. Because the universe is complex, increments in organismic complexity yield survival benefits: (1) more efficient extraction of energy and matter, (2) more flexible response to vicissitudes, (3) more effective search. J.J. Gibson's ecological psychology provides a clue to the advantages of sophisticated information processing while the lore of computational theory suggests that a complex computer is needed efficiently to perform complex computations (i.e. sophisticated information processing).

With cultural evolution new processes of thought appear. Abstraction is universal, but rationalization first appeared in ancient Greece, theorization in Renaissance Italy, and model building in twentieth-century Europe. These processes employ the methods of metaphor, metalingual definition, algorithm, and control, respectively. The intellectual and practical achievements of populations guided by the several processes and exploiting the different mechanisms differ so greatly as to warrant separation into cultural ranks. The fourth rank is not completely formed, while regions of the world and parts of every population continue to operate by the processes of earlier ranks. It is the continued emergence of fourth rank and even fifth rank cognition that constitutes the coming singularity, not the emergence of “superintelligent” computers, whatever they may be.

What economists have identified as stagnation over the last few decades can also be interpreted as the cost of continuing successful engagement with a complex world that is not set up to serve human interests. Two arguments: 1) The core argument holds that elasticity (ß) in the production function for economic growth is best interpreted as a function of the interaction between the economic entity (firm, industry, the economy as a whole) and particular aspects the larger world: physical scale in the case of semi-conductor development, biological organization in the case of drug discovery. 2) A larger argument interprets current stagnation as the shoulder of a growth curve in the evolution of culture through a succession of fundamental stages in underlying cognitive architecture. New stages develop over old through a process of reflective abstraction (Piaget) in which the mechanisms of earlier stages become objects for manipulation and deployment for the emerging stage.

We can think of the self as the result of interaction between subcortical systems for regulating the global brain state and largely cortical systems for representing the current body state and autobiography. The personal pronoun system is at the interface between the cortical and subcortical systems. By constructing a network model for the pronoun system that is grounded in basic machinery for social interaction we show how the pronoun system allows speakers to achieve self-reference and how this capacity engenders the illusion of a unified self. The same model allows us to see that there is no essential difference between reliving incidents from one’s own past and giving life to imaginary characters in ritual and in literary works. Such imaginative experience may play a role in maintaining the coherence of the self through different emotions.

I treat a single word, Xanadu, as a “meme” and follow it from a 17th century book, to a 19th century poem (Coleridge's "Kubla Khan"), into the 20th century where it was picked up by a classic movie ("Citizen Kane"), an ongoing software development project (Ted Nelson's Project Xanadu), and another movie and hit song, Olivia Newton-John’s Xanadu. The aggregate result can be seen when you google the word, you get 6 million hits. What is interesting about those hits is that, while some of them are directly related to Coleridge's poem, more seem to be related to Nelson's software project, Olivia Newton-John’s film and song, and (indirectly) to Welles' movie. Thus one cluster of Xanadu sites is high tech while another is about luxury and excess (and then there's the Manchester Swingers Club Xanadu).

A life-long intellectual journey, from an interest in Coleridge's "Kubla Khan" to an interest in large language models.

Coleridge's "Kubla Khan" has a very coherent structure. Two movements of the poem are each divided into three sections; in both cases the middle of those three in turn has three subsections and again, the middle of the middle has three subsections. The first movement ends with "A sunny pleasure-dome with caves of ice," a line which is then repeated at the structural midpoint of the second movement. This structure encompasses both semantics and sound, uniting both in a single coherent mental act. The semantics of the poem’s first movement involves a series of cognitive blends in which the neural self provides one input while Xanadu imagery provides the other. The semantics of the second movement involves manipulating the reality status of successive mental spaces. Underlying the entire poem is a “walk” by core brain mechanisms tracing territorial, sexual, and attachment patterns through the poem’s semantics. Coleridge’s 1816 preface embodies an abstract pattern that paradoxically asserts and denies the poem’s validity. On the internal evidence, the poem is whole and complete.

This is a review essay of Benny Shanon, The Antipodes of the Mind: Charting the Phenomenology of the Ayahuasca Experience (Oxford University Press, 2002), which is a detailed phenomenological account of what happens when one takes ayahuasca, a psychoactive concoction from South America. Ayahuasca is traditionally taken in the company of others and accompanied by music, which paces the visions and affects their content. The effect is that of entering another world. I offer some speculation about underlying neural processes, much based on the work of Walter Freeman’s speculation that consciousness is organized as short discontinuous whole-hemisphere states. I also speculate on the similarity between the dynamics of ayahuasca experience, as Shanon has described it, and the dynamics of skilled jazz improvisation; and I point out that what Shanon reports as a second-order vision seems to be involved in Coleridge’s “Kubla Khan”.

We can think of the self as the result of interaction between subcortical systems for regulating the global brain state and largely cortical systems for representing the current body state and autobiography. The personal pronoun system is at the interface between the cortical and subcortical systems. By constructing a network model for the pronoun system that is grounded in basic machinery for social interaction we show how the pronoun system allows speakers to achieve self-reference and how this capacity engenders the illusion of a unified self. The same model allows us to see that there is no essential difference between reliving incidents from one’s own past and giving life to imaginary characters in ritual and in literary works. Such imaginative experience may play a role in maintaining the coherence of the self through different emotions.