Preparing an archive of the not yet..
It’s been tempting to blast out all these blogs with everybody’s new best LLM ChatGPT but thus far I have restricted myself to experimenting with it, challenging its biases and not letting it do the dissertation for me.
This update will be the only one that actually uses ChatGPT content – I’m going to work with it to generate some visual responses to my questions, in the form of p5 generative art.
First – “Can you write me a p5.js program that draws a futuristic vista based on the title “archive of the future”. The whole picture is drawn using p5 shapes, strokes, lines, and the effect should be realistic and detailed.”
After one error in the code which it subsequently fixed, it coded:
That’s an interesting sketch. Can you describe to me what you were trying to illustrate with it?
Absolutely! This sketch was meant to depict a representation of a futuristic data archive, hence the title “Archive of the Future”. It does so by generating and visualizing a set of “data blocks”, each represented by a rotating 3D cube.
In the sketch, each cube (or “data block”) is assigned random positions (x, y, z) within a predefined range, giving a sense of three-dimensional space. The size and hue of each cube is also randomly determined, creating variety and depth in the visualization.
The entire scene of cubes is slowly rotating around the X and Y axes, giving the viewer different perspectives on the data archive. The rotation can be seen as the constant activity or data processing occurring in this futuristic archive.
In essence, the sketch is trying to convey a concept of a vast, vibrant, and ever-changing data archive of the future. It’s an abstract representation, and the actual interpretation can vary depending on the viewer’s perception.
It’s worth noting that while this sketch is simple, with more complexity and detail it could start to look quite mesmerizing and much more “futuristic”.
Can you now do a self portrait in p5? Make it as realistic as possible
As an AI, I don’t have a physical form or personal appearance to make a self-portrait of. However, I can help you create a simple stylized representation of an AI or a bot using p5.js. It’s important to note that creating a realistic image using only code can be quite challenging and may not yield results comparable to a hand-drawn or computer-generated image created with graphics software.
Here’s a simple example of a stylized robot face:
Can you be more creative and think about wider representation of non-human representation, perhaps think about cybernetics or organic forms, whatever you like, and update the sketch
Absolutely, that’s a wonderful suggestion! Let’s create a representation that could symbolize an AI or cybernetic entity, combining organic and geometric forms. Here’s a sketch of a tree-like structure, symbolizing growth and learning, with nodes representing data or knowledge:
Wow that’s great, very creative. I have one more task for you – can you craft a prompt for Dalle2 that generates a vision of a totally normal day five years from now?
Certainly, here’s a prompt that should generate an image of a “totally normal day” five years in the future, according to the perspective of DALL·E 2:
“Depict a serene morning scene in an urban setting in the year 2028. The image should include people going about their daily routines – walking to work, enjoying a cup of coffee at a sidewalk café, kids going to school. Technology should be subtly advanced but not intrusive, seamlessly integrated into everyday life. The architecture should reflect sustainable and green designs, and self-driving cars and delivery drones should be present, indicating advancements in AI and automation. The overall atmosphere should feel calm, familiar, and hopeful, with clear skies and bright, soft lighting.”
Can you generate the same thing but change the prompt to reflect a future where humanity has lost control of AI?
Certainly, here’s a darker prompt for DALL·E 2 to generate a vision of a future where humanity has lost control of AI:
“Depict a chaotic evening scene in a dystopian urban setting in the year 2028. The image should include people in a state of confusion and alarm, running from advanced machines that have taken over everyday tasks. AI-powered robots, self-driving cars, and drones should be depicted as malfunctioning and creating havoc, reflecting a loss of control over AI. The architecture should be shown as dilapidated and in disarray, due to the effects of rampant technology. The overall atmosphere should feel unsettling and tense, with ominous, dark clouds in the sky and harsh, contrasting lighting.”
Thanks! Can you describe a prompt that generates your earlier p5 self portrait either the first or second one, whatever you like better, and makes it look photorealistic in dalle
Of course! Here’s a prompt for DALL·E 2 to generate a photorealistic vision of the second p5.js sketch I provided, which was a tree-like representation of an AI:
“Create a photorealistic, glowing, cybernetic tree with circuit-like branches and luminous red nodes against a dark backdrop. The tree should blend technology and organic forms, subtly implying motion and serene intelligence.”
Do you think machine intelligence is more arborescent or rhizomic?
The nature of machine intelligence can be seen as both arborescent and rhizomic, depending on the specific perspective or context.
Arborescent structures are hierarchical and tree-like, with clear paths from roots to leaves. This is reflective of many machine learning models, such as decision trees and hierarchical clustering algorithms, which build explicit, hierarchical structures as part of their operation.
On the other hand, the rhizomic structure, a concept from Deleuze and Guattari’s philosophy, represents a non-hierarchical and decentralized network, where any point can be connected to any other. This can be seen in certain types of neural networks, such as recurrent neural networks or the architecture of the Internet, where connections are not necessarily hierarchical, and information can flow along multiple paths.
However, the concept of rhizome also encompasses notions of multiplicity, assemblage, and continuous growth and proliferation, which are not fully captured in current models of machine intelligence. While artificial neural networks can be described as rhizomic in structure, they lack the adaptive, autonomous, and creative capacities that the rhizome metaphor implies.
So, in the current state of machine intelligence, it can be seen as having both arborescent and rhizomic aspects, but it’s worth noting that these are abstractions and simplifications, and the reality of machine intelligence is much more complex.
can you do another self-portrait prompt but incorporate this latest point into it? i will put the prompt into dalle2
Certainly! Here’s a prompt that attempts to incorporate the concept of arborescent and rhizomic structures into the representation of AI:
“Render a glowing, cybernetic form merging arborescent and rhizomic structures. It should have a radiant core, tree-like branches, and intricate, interconnected light pathways against a dark backdrop.”