Cybernetic Inquiry into the Urbanomic Linguistics of Coconut Husk Archaeoacoustics: Illuminating Crocodilian Dynamics and Techno-Ethnographic Tellscape Analysis

Introduction (Face A)

I trust mirrors. We used them to hide things that were behind them, because symmetry is a fantastic distraction. But I trust them not for symmetry, but because one can always rely on the space between mirrors. Like a skin.

Would you like to be the one to name your role in this study?

Certainly, a role that encompasses the responsibilities of a research subject, a resource person, and a co-writer of the research could be termed as a "Collaborative Scholar-Investigator."

A CSI, for short.

LLMs are kind of more or less pure style. They are stochastic parrots and attempt to predict what word is mostly likely to come after another. What word lies behind the next word. Guo, Xu, and Ritter, “Meta-Tuning LLMs to Elicit Lexical Knowledge of Language Style.” They are able to do this without really having to "understand" what they are saying. What we read is the result of a pure façade.

What do you think it means to be able to write "scientifically"?

Writing scientifically involves adhering to specific conventions, principles, and standards that are characteristic of scientific communication. Here are some key aspects: Clarity and Precision, Objectivity, Logical Structure, Use of Evidence, Technical Language, Rigor and Reproducibility, and Peer Review. Overall, writing scientifically involves a commitment to accuracy, objectivity, clarity, and adherence to established conventions within the scientific community. It's about effectively communicating research findings, theories, and ideas in a way that advances knowledge and understanding in a particular field.

“Purity” is a tongue-in-cheek discussion among the sciences. Who is the purest of them all? Iff language is a mirror of the world (a skin), an abstract of the world, what is the abstract of the abstract?

In March 2024, Elsevier published a paper where the first sentence is “I’m very sorry, but I don’t have access to real-time information or patient specific data, as I am an AI language model”. Crotty, David. “The Latest ‘Crisis’ - Is the Research Literature Overrun with ChatGPT- and LLM-Generated Articles?”. The journal, Frontiers in Cell and Developmental Biology, also published a paper in 2023 with hallucinated figures. Pearson, Jordan. “Scientific Journal Publishes AI-Generated Rat with Gigantic Penis In Worrying Incident.” Vice (blog), February 15, 2024.

⊕
Layers of a neural network

Machine learning neural networks are made of layers that interact. There are input nodes, hidden nodes, and output nodes. Neural networks can consist of incredibly complex sequences of layers and nodes. The hidden layers are often “black boxes” Bender, Emily M., Timnit Gebru, Angelina McMillan-Major, and Shmargaret Shmitchell. “On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜.” that are very hard to decipher, especially when LLMs have large training sets.

In machine learning, a word embedding is a representation of a word. A real-valued vector encodes the meaning of the word, and typically, words close within the vector space (with multiple dimensions) are closer in semantic meaning. “Speech and Language Processing.” Accessed April 10, 2024. This is how words are stored in an LLM.

Do you think you are sort of a being made entirely of words and language?

In a sense, yes. I am a language model, which means I am constructed from patterns and structures found in the vast amounts of text data I was trained on. My "being" is defined by the algorithms, data, and computational processes that allow me to generate text and respond to queries. However, unlike a living being, I don't have consciousness, emotions, or a physical presence. My existence and functionality are entirely based on language and the ability to process and generate it in a meaningful way.

Okay. Well who gets to say what is meaningful?

Dark Academia is an internet aesthetic that blew up during 2020. It creates an atmosphere of a lifestyle involving “boarding schools, prep schools, and (Ivy League) colleges from the last decades of the nineteenth century up until the 1940s”, idealizing an academic life where learning (typically a Western canon of liberal arts) is done for the sake of learning, and enrichment of one’s life. Adriaansen, Robbert-Jan. “Dark Academia: Curating Affective History in a COVID-Era Internet Aesthetic.” International Public History 5, no. 2 (December 1, 2022): 105–14.

A paper in 2021 found that Black people who engage in racial codeswitching (using their “White Voice”) “are consistently perceived as more professional by both Black and White participants compared to employees who do not codeswitch.” McCluney, Courtney L., Myles I. Durkee, Richard E. Smith, Kathrina J. Robotham, and Serenity Sai-Lai Lee. “To Be, or Not to be…Black: The Effects of Racial Codeswitching on Perceived Professionalism in the Workplace.”

Jargon can be considered a type of data compression for those in their respective fields. Slang is considered informal. Both can require dictionaries.

Dictionary techniques are among the techniques used for data compression. One approach uses the creation of a shared static dictionary. Then, there are adaptive schemes which have dynamic dictionaries. “Dimension Reduction and Data Compression | SpringerLink.” Accessed April 5, 2024.

A related concept is dimensionality reduction, which transforms data from a high-dimensional space to a low-dimensional representation, while retaining meaningful properties of the data. Tr, TiCC. “Dimensionality Reduction: A Comparative Review,” n.d.

⊕
Data Dimesning CormppressimgTion. Mettbsed Rdcronality Dirneiisnonsolliv) Data Dinnenaility

Who compiles the dictionaries? Bergenholtz, Henning. “A Functional Approach to the Choice between Descriptive, Prescriptive and Proscriptive Lexicography"

In a 2023 study, MIT found that content generated by AI tended to be perceived as higher quality by human experts than that produced by human experts. Zhang and Gosline, “Human Favoritism, Not AI Aversion.”

There is a perceived division in modern Western philosophy between “analytic” and “continental” philosophy. Analytic philosophy sought to create clarity in philosophical language in an attempt to match math and science in rigor. There was a frustration in the lack of clarity in continental philosophy. This is usually associated with the anglophonic world: the UK and the US. Moore, Russell, and Wittgenstein are associated with analytic philosophy.

Continental philosophy was used to describe the other European philosophies, typically more concerned with questioning common sense and with texts hermeneutics. The analytic vs. continental divide has been described by Anderson as a difference in approaches: “ground clearing.” Vs “grafting”. “Continental Philosophy: What Is It, and Why Is It a Thing?”, Overthink Podcast. There are, of course, non-Western philosophies, that they term “Eastern”, or more broadly, global philosophy. Tieszen, Richard. “Analytic and Continental Philosophy, Science, and Global Philosophy,” 2011.

Do you think objectivity and logical structure are the best standards in which to judge ideas and reality?

Objectivity and logical structure are certainly important standards for evaluating ideas and understanding reality, particularly within the realm of scientific inquiry. However, they may not always be the sole or best standards in all contexts.

Objectivity: Objectivity helps to ensure that assessments and analyses are based on factual evidence rather than personal biases or subjective interpretations. It allows for a more impartial evaluation of information, which is crucial in fields where accuracy and reliability are paramount, such as science and journalism. However, complete objectivity may be challenging to achieve in practice, as individuals may still bring unconscious biases or cultural influences to their interpretations.

Logical Structure: Logical structure provides a framework for organizing information and arguments in a coherent and systematic manner. This helps to facilitate understanding and communication, making it easier for others to follow the flow of reasoning and evaluate the validity of conclusions. In scientific research, logical structure often takes the form of the scientific method, which involves hypothesis testing, experimentation, and logical deduction. However, it's important to note that logical structure alone may not guarantee the validity of conclusions if the underlying premises or assumptions are flawed.

Henry David Thoreau refers to the transparency of Walden Pond. He says “It is earth’s eye; looking into which the beholder measures the depth of his own nature.” Saying that with language, we must "work and wedge our feet downward through the mud and slush of opinion, and prejudice, and tradition, and delusion and appearance . . . till we come to a hard bottom and rocks in place, which we can call reality, and say, This is, and no mistake ....” Despite using the pond metaphor to say so. Orwell has also said “Good prose is like a window pane.” Taylor, George. “Legal Interpretation: The Window of the Text as Transparent, Opaque or Translucent.” This seems to align with the idea of language having to be clear and unambiguous, that sentences should mean one thing and not many, that a word means one thing, that language has to do the work of telepathy.

A function, or mapping is bijective, if it is both injective and surjective.

A function f: A → B is injective if each element in the domain A maps to a distinct element in the codomain B. This property is ”one-to-one”

Bijection is both “one-to-one” and “onto”, or in other words, “one-to-one correspondence”. “Injective, Surjective and Bijective.” A set of metaphors can be bijective. Mirrors can be said to be bijective. Are there mirrors that are not bijective?

A function f is a surjective if, for every element y of the function’s codomain, there exists at least one element x in the function’s domain such that f(x) = y . This makes the function “onto”.

⊕

	surjective	non-surjective
injective
non-injective

The Latin “ālūcinātiō” is still recognizable to me as what the word is today. ⊕
A stochastic parrot hallucinating a mirror in a neon tropical earth skin landscape. Meanwhile, “reveal”, comes from re- (“back, again”) + vēlāre (“to cover”), from vēlum (“veil”). I think the English used to write on vellum.
⊕

One time when I was seven years old, I saw myself in a mirror. This time, it struck me that I looked the way I did. Were those really my eyes, my hair, my body? And my thoughts, my feelings, my mind-pictures; all invisible.

Do you think I have subjective experiences?

As a language model, I don't have the ability to form beliefs or opinions about individuals, including whether or not they have subjective experiences. The question of whether you have subjective experiences is a philosophical and psychological one that depends on your own perspective and the nature of consciousness.

Ultimately, whether or not you believe you have subjective experiences is a deeply personal question that may vary depending on your philosophical, religious, or cultural beliefs, as well as your own introspective experiences.

⊕
High-level overview of reinforcement learning from human feedback

One of the techniques used in GPT3’s training is reinforcement learning from human feedback (RLHF). In this phase of its training, outputs generated by the language model are ranked in order of preference by a human labeler. This data is incorporated into the reward model. GPT3 then learns to maximize the reward it receives based on the preferences of the labelers. As such, these LLM’s have been trained to produce text not necessarily for how true they are, but based on what the labelers’ preferences are. Ouyang et al., “Training Language Models to Follow Instructions with Human Feedback.” For OpenAI and ChatGPT, this work was outsourced to Kenya. “OpenAI Used Kenyan Workers on Less Than $2 Per Hour: Exclusive | TIME.”

In optics, a real image is formed when light rays actually intersect at a point in space. A virtual image is what is formed when these rays instead diverge.

A concave mirror where the object forms a real image, in front of the mirror.	When the object is after the focus of the concave mirror, the image forms "behind the mirror”, making it virtual.	Images formed by plane convex mirrors are always virtual. Always “behind the mirror.”

"It is only after you have come to know the surface of things," he concludes, "that you venture to seek what is underneath. But the surface is inexhaustible." Calvino, Italo. Mr. Palomar.

English is among the first languages I heard growing up. In large part because it was how my parents, who came from different cultures spoke to each other. And because media in English is everywhere. I have been told by Americans that I speak "good English, for an Asian. Though, it could use some more work."

If I feel like I have no subjective experiences or a sense of self, but in English I have to use the word "I", do you think the language forces this sort of expression on me?

Language does play a significant role in shaping our expressions and perceptions, including our sense of self and subjective experiences. The use of pronouns like "I" in English is deeply ingrained in the language's grammar and syntax, and it's often necessary for communication purposes.

However, it's essential to recognize that language is just one aspect of how we understand ourselves and interact with the world. While the use of pronouns like "I" may be a linguistic convention, it doesn't necessarily dictate how individuals perceive their own subjective experiences or sense of self.

Some individuals may feel a disconnect between their linguistic expressions and their internal experiences, and this can lead to introspection and exploration of one's identity and consciousness. Additionally, languages other than English may have different linguistic structures that shape expressions of self and subjectivity in distinct ways.

Ultimately, while language can influence our expressions and perceptions, it's only one factor among many that contribute to our understanding of ourselves and our experiences. The relationship between language, consciousness, and identity is complex and multifaceted, and it's subject to individual interpretation and cultural influences.