“And if the earthly should forget you,
say to the silent loam: I flow.
To the rushing water speak: I am.”
“And if the earthly should forget you,
“And if the earthly should forget you,
say to the silent loam: I flow.
To the rushing water speak: I am.”
Or how I am spending too much time on ChatGPT. ChatGPT is now being incorporated into the Bing Search engine BTW.
And forget about the hand wringing about teachers being underpaid to read computer generated text or the oft-repeated end of the humanities, most dangerously, this will make it easier for people to write malware.
In terms of data literacy, literacy is literacy, right? So I asked ChatGPT to give me a simplified overview of what it does when a user types text into its interface. As an aside, I have been testing it with gibberish much akin to the Turkish thriller Hot Skull (see another post): communication (ergo civilization) depends on iterative translation of understandable context and content – the Chomskean reference in the first episode of the show, I assume.
As you can see below from ChatGPT’s response it is straight forward semantic analysis but this semiology is, to me, not the same as that for humans: machine semiotics ≠ human semiotics. But how is it different is what I am most interested by. Can machine dynamic semantics be trained solely using iterative utterance pairs? What can I learn about human learning from learning about machine learning trying to mimic human learning. Am I caught in a cyclical redundancy error?
When I receive a user's input, here is a more detailed explanation of the process I go through:
For example, I know that verbs are often used to describe actions or states of being, while nouns are used to name people, places, or things. Adjectives are used to describe nouns, and adverbs are used to describe verbs, adjectives, and other adverbs.
1. Tokenization: The first step is to split the input into individual words and punctuation marks, a process called tokenization. This allows me to analyze the input at the word level.
2. Part-of-speech tagging: Next, I assign a part of speech (such as noun, verb, adjective, etc.) to each word in the input, based on its definition and its role in the sentence.
3. Dependency parsing: I then analyze the relationships between the words in the input, taking into account their grammatical roles and the structure of the sentence. This allows me to understand the meaning of the sentence as a whole.
4. Sense disambiguation: If a word has multiple meanings, I use the context of the sentence and the rest of the input to determine which meaning is most appropriate.
5. Entity recognition: I also try to identify any named entities (proper nouns) in the input, such as people, places, and organizations.
6. Response generation: With all of this information, I can then use my knowledge and understanding of language to generate an appropriate response. This might involve looking up information, using my pre-defined responses to similar inputs, or using my understanding of the user's intentions to formulate a response.
In order to determine the role that a word is playing in a sentence (also known as its part of speech), I use my understanding of grammar and the relationships between words.
By analyzing the relationships between words in a sentence and their definitions, I can determine the part of speech of each word and understand the overall meaning of the sentence.
I followed up with a question about what item 4 was, which seems a key step in its assignation. How does it connect content with context?
Sense disambiguation is the process of identifying the correct meaning of a word in a given context. Words can often have multiple meanings, and it's important to choose the right one in order to understand the intended meaning of a sentence.
For example, the word "bass" can be a noun referring to a type of fish, or it can be a noun referring to a low-frequency sound or a range of notes in music. In order to disambiguate the sense of the word, it's necessary to consider the context in which it is used. If the word appears in a sentence about music, it is more likely to refer to a range of notes, while if it appears in a sentence about fishing, it is more likely to refer to a type of fish.
There are a number of techniques that I can use to disambiguate the sense of a word, including analyzing the surrounding words and phrases, considering the overall topic of the conversation, and using my general knowledge about the world. By using these techniques, I can choose the most appropriate meaning for a word and better understand the intended meaning of the sentence.
I asked ChatGPT is come up with the plot for a cop buddy movie between a person (me) and an AI program. While not going to win points for originality, I am really intrigued by the two “mishaps and misadventures”, i.e. fire alarm and the vending machine.
I’m also interested that I didn’t provide the name or gender of the AI but I wonder if the word “sassy” would have been used if the name chosen wasn’t a “feminine” name. This may be an example of learned bias being reinforced.
The movie follows the adventures of two unlikely partners: a bumbling police detective named Michael and a sassy artificial intelligence named Alice. When Michael is assigned to a high-profile case involving a string of robberies, he is paired with Alice, a newly developed AI designed to assist with investigations.
At first, Michael is skeptical of Alice's capabilities and unwilling to trust a machine to help with his work. But as they team up to track down the perpetrators, Michael starts to see that Alice is more than just a machine - she is a valuable partner with her own unique skills and insights.
As they work together to solve the case, Michael and Alice encounter a series of hilarious mishaps and misadventures, including accidentally setting off the building's fire alarm while trying to hack into a computer, and getting into a heated argument with a vending machine over the price of a snack.
Despite the challenges they face, Michael and Alice always manage to come out on top thanks to their combined wit and determination. In the end, they successfully crack the case and bring the perpetrators to justice, becoming the most unlikely crime-solving duo in the city. As they celebrate their victory, Michael realizes that he has learned to appreciate and trust Alice's unique abilities, and the two vow to continue working together to keep the city safe and laughing.
Each time I practice I look to one or more of the following thoughts – what do I confront? What do I transform? What do I integrate or innovate? I always like to have a clearly stated intention when I begin.
I intend to assemble here and in subsequent posts, accessible objects that I use in my daily ritual.
I begin each day during brahma muhurta, one hour and thirty-six minutes before the rise of the sun. This is the penultimate muhurta, a period of 48 minutes – 1/30th of a day. Sunrise in Ottawa today was at 7:38.
I will digress that this word muhurta in Sanskrit is made up of two sounds that refers to the immediate and to how this immediacy functions as an ordering – a perpetual now found in thinkers familiar in the Western tradition from Heraclitus to Nietzsche to Latour.
I will order my bed and brush my teeth before I begin my wake-up routine of pranayama and meditation. I was told a long time ago that monks always start with ordering their space in preparation for practice which makes a lot of sense to me even though I am no monk. And while my meditation always begins during brahma muhurta, my pranayama may end after sunrise.
My goto app in the morning for meditation is Oak. I dont use it any other time except for my morning routine. It is a habit. Its free and does exactly what I needed it to do when I started this ritual when I was at home during the pandemic. It show my progress (good for motivation since I easily depreciate my accomplishments) and has both breathing and mediation timers including box breathing. I dont’ use it for Tummo sessions which I will describe in another post.
Oak is pretty basic but a nice GUI and breathing shouldn’t be anything but basic anyway. IOS only. I did pay for the course and it was worth the 2 bucks I paid, if, for nothing else, as an offering to initiate my ritual.
*Breaking the chain refers to Jerry Seinfeld's rule about being successful at things - using a visual indication - in his case a calendar where he puts a big X each day he writes jokes. By having a visual cue, it was much easier for me pay attention to creating rituals and habits. It is ritual, according to Oscar Wilde, that is the origin of religion and, for us moderns, the basis of Cognitive Behavioural Therapy too.
I was using pillows on the floor or a chair and decided to upgrade my meditation setup during the pandemic. Again, like breathing, this should be simple. I started on a chair and laying down. And even my zabuton from Half Moon, while comfortable, isnt perfect but all you really do just need to sit. Or lie down. Pretty simple. Don’t buy anything, all you need is your breath.
I still enjoy and watch this series : 7 Days to Mindfulness with Kirat Randhawa. I appreciate her calm demeanour in this series along with some guidance on topics such as finding your centre and how clarity and curiosity work in the meditation process and its use in day to day life when I am not in mediation. I also watch Netflix’s Headspace Guide to Mediation which was very accessbile and enjoyable. It was a great beginner resource for me when I initially found it, a perpetual beginner.
I come back to these teachings once in a while and find new things in them, each time! Next up, my pranayama practice and books!
A lot of people are talking about the new algorithmic “chat bot” that will, according to some, herald the end of the humanities as we know them. And while I have found some super interesting things that the chat-bot can do with “simple” prompts, it is quite interesting to reverse engineer some things that you might take for granted. More to come on this as I tinker.
The link that I will suggest is that StackOverFlow has added a new, temporary, policy directly about the use of Chat GPT stating that, among other things, the requirement for citations and the ability of the community to judge how accurate the answer responded to the query posted is core to this action. It is this latter part that is most interesting to me: persons with subject matter expertise may interpret the initial question (stimulus) and suggest what the requestor really wanted to know with this line of inquiry. See Metaphoric and Metonymic knowledge by Roman Jacobson and his influence on Lacan and Barthes. Chomsky too, for a delicious connection with Hot Skull below.
And while I tend to avoid Canadian politics, I will suggest that the fact that Canadian business people actively sought and received public COVID aid and then didn’t spend it for its intended purpose of employment is pretty fucked up. I appreciate this type of reporting from Jacobin.
I finished two television series this week, both well worth the time. The first is Mike White’s White Lotus, season 2. Review on Jacobin here that I found after reading above mentioned article about corruption. White Lotus is filled with lots of great characters and the scenery in Sicily was awesome with Mount Aetna in the background as in my screenshot below. I am excited for season 3. And hope to see more of Jack from Essex who, like the son in Season One, remains hopeful although broken. Will Greg make an appearance in season 3? Portia? I’m eagerly awaiting its arrival!
As I explained to a colleague, what I loved the most about White Lotus was that the initial episode (the beginning) was the end of the story but I was constantly amazed at how things that I assumed were true as the story progressed only appeared that way because of how I initially interpreted key scenes from that first episode. Great storytelling that showed many of my own biases and affectations with various herrings strewn about. It certainly kept me entertained.
And TIL that the writer, Mike White, was on the US reality tv show Survivor.
The second series is Hot Skull from Netflix Turkey. I loved the back drop of Istanbul. A virus spread by speech that drives people into confusion and stupor? Certainly intriguing. It was in the first or second episode where they mention Noam Chomsky and fitting this pandemic along with an ambitious leader exploiting the crises for his own self aggrandizement is a recipe for success.
I spend a lot of time online, as many do. Even this website is mostly just used for my own purposes, the WordPress front to this site is a vanitas project that I mostly used while I travelled but who travels anymore. I still code in the background but have been struggling to make posting a habit.
To that end, perhaps I will add links to stories that I enjoyed reading. I told myself that this was a great way to collect data on what I am paying attention to. That was enough to motivate me. I think. I’ll start tagging posts too that will help me create metrics for the data collection exercise.
First up, China. I have many unpublished writings on the geopolitics of China and Ukraine, mostly related to the historiography of Diplomatic History. My draft folder is replete with stories about the decline of empire but as I incorporate more and more history back into my “Cold War” reading list, I realize how important writers such as Odd Arne Westad are still changing my thinking about these issues as much as Melvyn Leffler did in grad school.
This Wired article about the protests in China is the most important piece from this week. COVID mobility restrictions is increasingly oppressive in many areas as officials struggle to immunize this vast population. Couple this with the inability of Xi to take full advantage of the transition of Hong Kong and the death of Jiang Zemin, and it is no wonder that the ghosts of Tiananmen are on the mind. Funerals and earthquakes tend to be external events that motivate social change in China.
The vector to technology and the Great Firewall of China is important. PRC have loosened mobility restrictions in Guangdong province but doing this may fuel discontent in other regions as they learn of this. Can central authorities maintain message control in the light of this “white paper” revolution?
This book on Basquiat is the book release of the week (I type this like it is a thing), but art history remains central to my reading and how I spend (too much of) my time. I am working on an article on Philip Guston and one thing that I am constantly reminded of in his late work is the street graffiti of New York. And while Basquiat is normally seen through the eyes of Warhol, Guston may be more appropriate in an asemic, object oriented, perspective. I’ll write more when I get this book. It will go on the list.
So two links for this week. More to come.
I settle my body like a tree stump.
I hold my arm like the branch of a withered tree; out of the vastness of heaven and earth, the multitude of the myriad things, I am only aware of cicada wings.
I don’t wander or waver, and I would not take all the myriad things in exchange for the wings of the cicada.
How could it be that I’d fail to succeed?
Just thinking out loud.
The digital age means that even the tools of Benjamin’s mechanical reproduction are not representations of reality. For Benjamin, reality has a sense of being the present, an aura of presentness.
The role of the audience or the spectator, like in the Parisian arcades (proto Consumer), becomes part of the assemblage through bracketing.
Bracketing is like purchasing an object in capitalism- an act of consumption. They are transactional but is this Deleuzian?
The bracket of the arcaded city mirrors the bracket of the arcaded nation – the closing of the frontier heralded by Frederick Jackson Turner.
Immersive digital forms are not assemblages of photographs; rather they historical objects that are aesthetically curated – they can be bracketed.
Amidst all the other things going on I was pleased to spend time at the National Gallery yesterday. The new exhibition featured some exceptional works of arts from a range of Canadian artists from Emily Carr to James Wilson Morrice. A catalog was produced and, as usual, it is a well produced book with nicely coloured plates and good paper for viewing them. I will go several times to the exhibit after reviewing the catalogue and reading the essays.
For now my head is brimming with questions such as how these artists thought about impressionism and their pictured responses – and would the artists assembled in this collection think of themselves as impressionists, whatever that might have meant to each of them? The catalogue and its essays will help me understand this better from a curatorial and academic perspective. I am cognizant of the use of the term “impressionism” and how it was originally a satirical take on what once critic assumed was an “unfinished work”.
I take this as my starting point: as the camera aesthetic emerged as a means of visualization and “freezing time” (think Muybridge) and “documenting” time (think of the use of the camera for policing and anthropological itemization), imaginative works of paint were not limited by the “instant” nature of time and could allow interactive lighting effects between the viewer and the object of art to mimic time but non-synchronically.
As an aside I find that I need a multi-media or a multi-modal perspective, not a “virtual collage” but rather an assemblage of techniques and tools to reflect upon. Since each medium* has its own “perspective” and each has their own benefits, when I approach objects like this exhibition I want to create my own first impression before I consume too much of other people’s perspectives and biases accrete. Much like how I normally don’t like to listen to the audio guides the first time: I prefer to take my time and explore on my own initially. This is an entitled view, of course, since I can visit this Gallery frequently.
*Marshall McLuhan has been on my reading list of late. I have recently found a copy of the first article that I ever wrote on the internet back in the 1990s on McLuhan, filled with my embarrassingly youthful utopian ideals about the global village and its hope.
My interest in Impressionism is its treatment of light. I see this predominately in relation to the emergence of photographic tools and aesthetics and how these challenges were faced by painters.
Of course by the time I saw the first work at the Gallery, it was the colouring that most grabbed my attention. And notwithstanding some lighting issues in several of the rooms (since these rooms were designed for prints and not framed canvas works) with shadows that interfered with the artists composition (the Carr landscape has its yellow sky darkened), the works are a cornucopia of styles and techniques that provided several hours of viewing pleasure. The application of paint and its control was intriguing like this example, from Ernest Lawson’s Canal Scene in Winter.
I have been reading the essays from the catalogue and am struck by a remark that artists such as Maurice Cullen shared certain attributes with Hokusai. It struck a cord with me since I have shared that opinion not only of Cullen but that many works of easel paintings of landscapes from the mid-19th century share not just the colouring and loose brushwork but also that same “angle of view” that resembles an aerial perspective or a hilltop view as opposed to a eye-line frontal viewing perspective. Visualization in late Meiji (called “modernizing” Japan in the West) Japan shares many of these elements. This presages the aerial views that were popularized by painted visualizations from hilltops, tall buildings, airplanes or balloons in the early 20th century. I think of it as an “anthropological” perspective. It differs for me from a “cartographic” perspective used by the Dutch in Elizabeth Sutton’s work, for example. It assumes an objective rather than a notational perspective. It is this assumption of “objective” that is soon to be the subject of much attention in the first half of the 20th century.
The artist and the viewer have a larger focal length and more encompassing view that someone that would be a participant in the visual would have; not necessarily omnipotent but detached. So while the more “standard” impressionist work from the urban streets of Paris were photographic, many took the imagined perspective from slightly above. Note the Harris below compared with the Milne. Also, its opposite in the Henri Beau.
And my three favourite works from this visit are below. I cannot help but make the connection between the urban and industrial works and Burtynsky’s photographs in his Anthropocene.