Posts

Posted on

The Last of Us, according to AI

I’m using ChatGPT to write posts but find that the prompt engineering, like in midJourney, takes time to understand. It’s almost as much work to write it myself but here we are.

Where ChatGPT excels is listing stuff. I asked for some foreign examples in the post-apocalyptic genre from Europe 1900-1970.

“La Nuit des Temps” (Ice People) by René Barjavel (1968) – This French novel is set in a future world that has been devastated by a new ice age, and follows the story of a group of survivors who discover a way to travel back in time.

“Niekas nenorejo mirti” (Nobody Wanted to Die) by Viiljamas A. Padvejas (1962) – A Lithuanian science fiction novel about a post-apocalyptic world in which the majority of the human population has been wiped out by a mysterious disease, and a small group of survivors must find a way to rebuild.

“Alraune” by Hanns Heinz Ewers (1911) – A German science fiction novel set in a future world in which a war has wiped out most of humanity, and the remaining survivors must find a way to rebuild and restore order.

Hold the mushrooms.

The world of post-apocalyptic storytelling has been a source of fascination for many for decades. From the pages of books to the screens of films and TV shows, tales of survival and hope in a world ravaged by disaster have captivated audiences. One recent addition to this genre that has left a lasting impression is the television series “The Last of Us.”

Set in a world devastated by a fungal pandemic, the series follows the journey of Joel and Ellie, two unlikely allies on a mission to cross the country and find a resistance group known as the Fireflies. Their journey is a heart-wrenching tale of survival and human connection in a world where life as we know it has been forever altered.

What sets “The Last of Us” apart from other post-apocalyptic stories is its unique setting and the mastery of its storytelling. The origin of the pandemic sets the series apart, allowing for a focus on the science behind the collapse of civilization. The world of “The Last of Us” is brought to life through expertly crafted environments, from the overgrown ruins of cities to the lush forests, creating a sense of place that is both desolate and alive. The game’s sound design, which effectively uses environmental sounds to create a feeling of danger or serenity, further reinforces this sense of place.

The relationship between Joel and Ellie is the beating heart of the series, and their journey together is both touching and intense. The writing is top-notch, with well-developed characters and a well-paced narrative that keeps the viewer engaged from start to finish. The series’ use of cinematics is also noteworthy, as they effectively convey emotional moments and provide insight into the characters’ motivations and relationships. The voice acting, particularly the performances of Troy Baker as Joel and Ashley Johnson as Ellie, adds depth to the characters and brings the story to life.

In terms of visual imagery, “The Last of Us” is a feast for the eyes. The series’ use of light and color effectively sets the tone for each scene and underscores the emotional weight of the story. From the golden hues of a peaceful sunset to the sickly greens of a quarantine zone, the visual imagery of “The Last of Us” immerses the viewer in a world unlike any other.

“The Last of Us” is a standout in the post-apocalyptic genre, offering a unique take on a familiar theme. Its mastery of storytelling, expert world-building, and stunning visual imagery make it a must-watch for fans of the genre and anyone looking for an emotionally gripping tale of survival and hope.

Posted on

Yes, but is it art?

I was out for dinner with friends and discussed AI including AI and art and whether or not it is “art.” It was interesting since most arguments boil down to the fact that there isn’t enough human intervention!

I made this comic this morning by asking for a single frame comic in the style of The New Yorker then added text in Affinity Photo since MidJouney is really bad at text.

Posted on

ChatGPT and MidJourney

Where did the day go? Oh, right. ChatGPT can create some really good prompts for MidJourney when set up the right way. Here are a few of my favourites from today. I want to learn how to do better landscapes. But I am loving the schematics and blueprints.

Snow globe with Japanese temple.
I think this was luminous golden globe with intricate details lit from below.
Steampunk Schematic of the Eiffel Tower
Isometric cityscape
Posted on

MidJourney… wow!

Diving into AI with ChatGPT is bad enough but midjourney really impressed me and kept me up into the small hours last night. Here are a few renderings. I will add the prompt below the images. I have been using ChatGPT to build prompts but simple declarative statements deliver some great results.

Speaking of ChatGPT, here is a good article in The Atlantic on the potential that ChatGPT will put white collar workers out of work. I will find a research paper that I worked on years ago looking at an Oxford study on the impact of AI on labour markets and link it here with my other older documents. I converted their probability tables (based on SOC occupations used in the UK) into NOC to see the potential impact on recent immigrants to Canada – spoiler alert – bad for telemarketers, not bad for drug rehabilitation counsellors.

Paper cut out inspired by Dante’s Inferno
Celtic Raven tattoo
Meditating on the nature of time under a tree floating in space
And this one just still has me shook. It was my very first image and I asked it to paint Cy Twombly’s peonies in the late style of Philip Guston. The shadow is what really amazed me as those who have read my essay will know.
Posted on

First Binge of 2023: Severance

Severance Season 1

I was looking for a good series to binge after White Lotus and someone at my gym suggested Severance. I got a real “Lost” vibe from the show right from the start, some behind-the-scenes mysterious group controlling both the characters and us, the viewers – a common trope in contemporary drama. What are these employees doing? What is really happening? The Jim Carry movie Eternal Sunshine of the Spotless Mind was top of mind as well: how removing individual memories unravels the consistency that is the hallmark of the Self.

First off, as someone who appreciates cinematography, some of the location shots were pretty spectacular, the parking lot and the building interior were quite beautiful, allowing the architecture of the public spaces freedom as opposed to the claustrophobic spaces of the offices and the hallways that the workers inhabit. Like that hallway to the break room that John Turturo paints, the tightness of the working space is palpable.

I didn’t know that Ben Stiller directed but I noted he directed most of the first season. The characters unfold through the story until each revelation adjusts your understanding of what is “really” going on: whether cleaning the ocean, a more mundane experiment by a ego fueled billionaire, anonymizing state sanctioned killing, or some other mysterious earth saving initiatives (that was one of the connections that I had with the cultish Dharma Initiative on Lost), I am sure that we won’t ever know.

And malicious because of the hidden aspect of the work. Much like how the Milgram shock experiments showed the corporate banality of evil expressed by Hannah Arendt, hiding what each individual does, even to themselves, allows evil. Abstracting humans into data is bad enough.

I also didn’t know that Christopher Walken was in this so a big surprise when he entered the frame. His work at Lumen is as bizarre as the others and his presence adds even more weight to each scene that he inhabits.

What is up with that brother in law? Why do I get the impression that he is more behind this cult then we are led to believe? He reminds me of Patrick Swayze’s character in Donnie Darko. And why do I keep thinking about his explanation for the three beds in the kids room and its connection with the process of severance itself?

Not a lot of traffic about this show on Reddit but I did learn that they started filming season 2 a few months back so I eagerly anticipate its launch. The Guardian has a good season 1 wrap up from February 2022 here.

Posted on

Time is a tool that we use to create analogies between objects.

How we do this is by the art of comparison and proportioning. Analogies and metaphors are inferences and arguments about objects, necessarily phenomenological. It is nonsense to say that this object is different from that object but is the same. Aristotelian categorization and mathematical Platonism, let alone a Newtonian imperative and Schrondinger’s cat, demand that difference be expressed. 

As a thought experiment: if there were only 2 things in the universe, their difference can only be explained in terms of the other.

You create other objects with these parts of speech : bracketing two objects, one (or both) of which can be the self. But you cannot use these parts of speech to define an object itself; rather, you use these things to magically cast objects into being the same way that you use other types of figurative language such as antithesis, hyperbole, metonymy, or simile. Your use case may be deductive, inductive, exploratory, didactic, revelatory, obscurant or humorous. You may express this in rhetoric or painting or song and your success may or may not be as robust as desired by both you or your audiences.

In the fractal of the present, the time of creation, then, is the splendid feeling of Goethe. And while Goethe argued that “modernity” was destroying this sense of the “now” or the “present”, it can be argued that it is only an arrogance of persistence that does so and once freed from this constraint, harmony is restored.

This arrogance of persistence is the way that I see Tristan Gracia’s substantive challenge: seeing objects before we figure out the variables that we (as humans) are using to grasp it. This seems to be simply affects and aesthetics, concerned with how things are understood. But without vectors since objects have vectors, not Tools. Hence the issue with deploying Garcia’s thinking in a world of literal objects.

Posted on

More ChatGPT: learning about how to learn about er, learning.

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.
Posted on

ChatGPT miscellany

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.
Posted on

Pranayama and Meditation resources – One

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.

My current streak. Its on my old phone too so I always bring it with me when I travel since even though the app is on my new devices, my streak record is only on this phone and I cannot break the chain!*
*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!