New Thoughts on Old Thoughts
Perhaps GPTs are Generative Plastic Thought machines?
My simple thoughts about AI and the subgenre of LLMs in particular:
- GPTs are tools for extending your thinking flow and extending the kinds of thoughts you have.
- GPTs open the computer up to different kind of human interaction flow.
- Neither animal intelligence nor computing-machine intelligence will ever settle into answers and solutions. (I prefer to consider all intelligence phenomena as complex adaptation to consequences aka empathy of the situation.)
- GPTs give us all a new “unit of cultural production” that goes beyond the book, the music recording, the poem, the algorithm, the program, etc. We have these new “units of GPT”. (Wolfram calls them “computational essays”). GPTs do not replace any previous unit of culture. All additive.
Read My Thoughts
Everything in this post can be inspected, recreated, remixed, rethought here:
“Symbol and Relation, p5 thought machine”
Instructions(for those afraid of breaking thoughts):
- click and drag a thought shape anywhere even onto other shapes
- read and listen and think
- repeat
- click in upper right to get the editor and look at code etc.
- refresh browser whenever to start over
All of this is based on my old 2015 project of composing 500 philosophic thoughts in the order I had the thoughts. This is a philosophy, art and technology project I renew all the time with whatever the latest in tech and my skill and my free spactime allows.
Use Symbol and Relation GPT in your own ways
https://chat.openai.com/g/g-pjJm8zDEP-symbol-and-relation
Evolution of this New Thought Project:
Let's make a p5.js interface for Symbol and Relation.
1) assign a shape to each Thought ID based on it's string length and information entropy. (map length to some sort of polygon feature and the entropy to color gradient)
2) fill the canvas (1200x800 px) with key symbols from the work
3) provide a way for the user to type in 5 Thought IDs and then have those Thought IDs appear on the canvas. User can then drag the Thought ID shapes into each other or apart from eachother. when brought together combine the words of the thoughts by riffling the words of each thought together. and when Thought Shapes are seperated remove the thought from the combination.
4) consider adding simple sounds for combination and deconstruction
5) add other flourishes to the design, animation or interface as you see fit.
The Evolution of Thought
// Pseudo-code for the p5.js project
function setup() {
createCanvas(1200, 800);
// Initialize canvas, symbols, and key elements
}
function draw() {
background(255);
// Handle drawing symbols and interactive Thought ID shapes
}
function calculateEntropy(string) {
// Function to calculate information entropy of a Thought ID
}
function createThoughtShape(thoughtID) {
// Function to create a shape based on Thought ID characteristics
}
function mouseDragged() {
// Handle dragging of Thought ID shapes
}
function combineThoughts(thought1, thought2) {
// Function to combine thoughts when shapes are brought together
}
function separateThoughts(thought1, thought2) {
// Function to separate thoughts when shapes are moved apart
}
// Additional functions for sound effects, animations, etc.
function addNearestNeighborThought() {
if (Object.keys(thoughts).length === 0) return;
let lastThought = thoughts[Object.keys(thoughts)[Object.keys(thoughts).length - 1]];
let nearestThoughtID = getNearestNeighborThoughtID(lastThought);
if (nearestThoughtID) {
createThoughtShape(nearestThoughtID, thoughtIDs[nearestThoughtID]);
}
}
function getNearestNeighborThoughtID(lastThought) {
let minDistance = Infinity;
let nearestThoughtID = null;
for (let id in thoughtIDs) {
if (!thoughts[id]) { // Skip already existing thoughts
let distance = calculateDistance(lastThought.text, thoughtIDs[id]);
if (distance < minDistance) {
minDistance = distance;
nearestThoughtID = id;
}
}
}
return nearestThoughtID;
}
function calculateDistance(text1, text2) {
// Simplified distance measure between two texts (e.g., based on word overlap)
// ... Implement your logic here ...
}
So what was I thinking? What do I think now? What are you thinking?
Now, of course, there’s a deeper question to me in all that I did, my GPT did, what you might have done with it all.
Were all those thoughts I had in 2015 thoughts at all?
Are they still thoughts? the same thoughts? evolving thoughts? memories? imaginations?
and now are they programs? were they already programs?
who or what had any of these new thoughts? GPT? me? GPT and me? you now?
You can see that I clearly was responding the computer network and the computer network to me. And then in the interface it’s a new computer network and a new audience.
are the visuals part of thoughts now? are those new thoughts or “just a new frame”?
Have you ever attempted to map your thoughts out, computationalize them over and over and see if any of those thoughts still make sense?
Did you get to the bottom of this? Did you try the GPT?
I’ll know if you did because another set of computer networks are observing and will send me the stats of readership of this and any clicks on any of the other stuff.
But i’ll never know what you actually thought, will I? will I ever know what I think? what the GPT thinks? or if thinking is thinking and if thoughts ever settle down into A Thought.