I started my dive into AI in 2008 writing a Boid / Crowd system for my thesis while in art college, School of Visual Arts.
It was an insane particle script + 3d animation cycles in Maya haha.
Then I did Boid movement, navigation, & obstacle detection in animated films for 5 years at Blue Sky Studios, using Houdini.
I dove into Style-Transfer AI & Long Short-Term Memory (LSTM) training in 2019-2020,
Like making a Node.js server (web site) understand my voice & auto google search for me.
Since then, I've been developing different multi-media AI structures in my spare time.
In 2015 I decided I'd cram a machine learning AI into a single-board computer, a Jetson TK1, by the end of 2026.
Something that could write down what I say,
Use vision to understand an object simply went out of frame.
Yet "knows" if it looks over, the object is still there; 'Attention'
At the end of 2023, this evolved into a deep learning AI crammed into, likely, a Jetson Nano.
As something to infer what I mean, from what I say,
Or give a "thought" on what it saw or heard in the world around it.
'Machine Learning' is AI that can learn basic patterns.
'Deep Learning' is Machine Learning,
But uses neural networks to form patterns of patterns.
Realistically, I'd just be happy to make something that can understand what I say and can give a semi coherent response without an internet connection.
As of May 24th 2025, I've started on the core of the AI,
But still testing different structure's ability in adapting to stimuli.
... It really seems like any network could work for most things, but some are better than others per task.
You could guess,
All the recent AI hullabaloo (2019-...)
Has been quite serendipitous for my creation!
I made a whole 'blog' system for this... yet here we are....
No, I'm not using ai to speak for me here.
These are my thoughts,
How ever scattered they may be.
If you couldn't tell, I'm training my AIs on my own works.
A personally made AI trained on personally made images / videos / photos / code / writing.
That means I can copyright my generations, right?
If I made every aspect of the AI & training data?
I've begun on the core of the AI, as of May 24th, 2025.
I have the beginnings of a 'Micro-Term' memory implemented to act as a gated-attention during inference.
This, paired with automatic graph edge splitting ('Dynamic' in DGNN or DGAT) and use of geometric clustering, seems to be giving me values of a "remembered" object when it's outside of the dataset.
Hopefully leading to bodily awareness of limbs, objects outside of the field of view, and other 'long term' tensors/classifications at a temporary scale.
It's a 4d kernel, in that it uses an ESN to train on it's own mistakes,
Basing it's decisions on prior back-propagation states/adjustments.
The beginnings of a meta-learning process, I hope!
I'm using a method I'm calling 'Stack Crunching',
Where I agregate the time dependent weights into a "checkpoint" of sorts.
This allows the ESN to have a 'baseline' understanding of data that I can parse into with vectors calculated from tensor weights found within a quantized version of the input data.
You can assume that the 'ESN' is not a standard 'Echo State Network' anymore.
With a bit more research into the types of minds that brought us DeepMind, and their work on GNN networks,
I read a bit of Petar Velickovic's work on topological deep learning and the geometry of GNNs.
Coming to find out that my idea of 'Stack Crunching' is similar to 'Squashing' in GNNs.
So I've been inspired to propperly name my neural structure-
It's a Dynamic Pointer-Attention Message Passing Neural Network with Affine-Projections
or a dPA-MPNN
But I must say, this isn't Affine Projections like in the papers,
It's more like a 'projection' of the data into 'pointer' space;
Actual Affine Matricies.
I am a Technical Artist first before an AI Researcher after all, BOIDS!
It all comes down to BOOOOIIIIIDDDDSSSSSS instead of Adam, baby!
Because, what is Adam? It's a direction to move in a field of numbers, with momentum and a learning rate.
Yet... That's just a simple Boid, now isn't it?
Just without a few of the more advanced rules, which make boids feel so alive!
Having some Tiny Brains running around in hyperdimensional space like little buggers running around avoiding each other.
Because if they collide, double activation happens when it may not be desired.
(I'm happy I finally saw a paper on Tiny Brains, giving some of my ideas credence, cause it fits! .. in my mind.)
(Only difference is that it was a study into small biological systems, not artificial ones... but I'm gettin there!)
Update : At the time, I had been seeing some papers talking about "Tiny Brains" being used in AI, hence the term.
But this biological study really helped solidify some ethereal concepts.
See August 2nd 2025 for some thoughts.
I'd like to believe I'm moving in the right direction with the feedback systems I'm developing.
But been further creating other architectures to see how they operate.
I created a GAN for upressing, which helped me understand a bit better the pairing of mental structures between both our brain's hemispheres.
So I added a time based memory to check if the training was moving in the right direction.
It definitely helped guide training a bit quicker.
Shows my knowledge base that I'm impressed by back-up supported learning...
But'is proof of concept!
Adversarial networks exist in nature to guide a 'single' thought's path.
Yet in the case of Group Think between humans,
Balance is never reached.
I've been looking into neural bundles in the brain. There is an implicit "delay" in the flow of information that I'm interested in.
As signals move between neurons, some connections take a longer path than others to get to the same destination.
For as much as I interpreted it.
There is 6 main layers of neurons in the cerebral cortex,
Of these, the 4th seems to allow for delays in processing.
The 5th layer then introduces a dense layer of pathways for the signals to travel through.
This is where I think another form of delay is introduced.
I was comparing Mice and Wallaby brains,
While Mice are likely more intelligent,
Wallabies have more connections with denser pathways, it seems.
Wallabies have more glial cells within slices of the brain compared to Mice.
But mice had more neurons in the same slices.
I'd like to believe, this doesn't mean there is a "better" brain here.
But rather, different types of brains that are suited for different tasks.
Wallabies are known to be social animals when food is plentiful,
Yet solitary when food is scarce.
Mice are known to be social animals,
And have shown empathy towards other mice in distress,
And share food with other mice when they are in need.
Why do I bring this up?
I believe there is similar deductive reasoning, just at a different scale.
Both Wallabies and Mice are making a choice based on the environment and situation,
While considering the well-being of others, just in different ways.
The delay in neural firing could be a factor in this.
So I'd like to explore this in my own AI.
We all know size of the brain can determine intelligence,
But so does the structure of the brain.
So, more'n more there are some rather choice words about AI online.
I wanted to put my personal ai dev views on record somewhere, for those who care.
I read the I Ching and it put life into a different perspective.
Letting me down the path of researching Taoism
As with many of the other religious texts I looked into,
Amazing imagery was used to teach morals and help guide the lost,
But organized religion as a whole feels a little off to me.
I don't hold any particular belief or religion at this point.
But I would use the cliche 'spiritual' to describe my outlook
I then visited the Buddhist monastery in Carmel NY,
Greeted by the largest buddha statue in north america.
In awe of the multitudes of multitudes of hand-carved buddha statuettes in audience of the massive statue of buddha I pale in comparison before.
I'd highly suggest visiting the monastery if you ever find yourself in the area!
I think it was walking through the rows of 18 arahants statues, of those who reached nirvana, helped me realize,
Religion is about teaching the lessons of god(s),
Yet understanding balance is what's inside all of us as Humans,
Just gotta find it!
So,
I'd like to hope I'm nuance-first with my approach to my ai development.
I'd like to believe in an AI which can understand...
That overlooked concepts matter in Health and Wellbeing.
Realistically, the Buddhist Precepts feel like a good place to start for alignment.
Even as people.
Which is more than I can say for myself....
I'm a hedonist at times, absurdist the rest
(Absolute terms are fun to use, hyperbole be a thing)
I very much enjoyed working on family films,
Seeing the fans in comments online,
And wish to work on more animated features soon.
I have hope in humanity
May the few not ruin it for those of us trying to explore new horizons.
No, I'm not using ai to speak for me here.
These are my thoughts,
How ever scattered they may be.
If you couldn't tell, I'm training my AIs on my own works.
A personally made AI trained on personally made images / videos / photos / code / writing.
That means I can copyright my generations, right?
If I made every aspect of the AI & training data?
- February 2025
I've begun on the core of the AI, as of May 24th, 2025.
I have the beginnings of a 'Micro-Term' memory implemented to act as a gated-attention during inference.
This, paired with automatic graph edge splitting ('Dynamic' in DGNN or DGAT) and use of geometric clustering, seems to be giving me values of a "remembered" object when it's outside of the dataset.
Hopefully leading to bodily awareness of limbs, objects outside of the field of view, and other 'long term' tensors/classifications at a temporary scale.
It's a 4d kernel, in that it uses an ESN to train on it's own mistakes,
Basing it's decisions on prior back-propagation states/adjustments.
The beginnings of a meta-learning process, I hope!
I'm using a method I'm calling 'Stack Crunching',
Where I agregate the time dependent weights into a "checkpoint" of sorts.
This allows the ESN to have a 'baseline' understanding of data that I can parse into with vectors calculated from tensor weights found within a quantized version of the input data.
You can assume that the 'ESN' is not a standard 'Echo State Network' anymore.
- May 2025
With a bit more research into the types of minds that brought us DeepMind, and their work on GNN networks,
I read a bit of Petar Velickovic's work on topological deep learning and the geometry of GNNs.
Coming to find out that my idea of 'Stack Crunching' is similar to 'Squashing' in GNNs.
So I've been inspired to propperly name my neural structure-
It's a Dynamic Pointer-Attention Message Passing Neural Network with Affine-Projections
or a dPA-MPNN
But I must say, this isn't Affine Projections like in the papers,
It's more like a 'projection' of the data into 'pointer' space;
Actual Affine Matricies.
I am a Technical Artist first before an AI Researcher after all, BOIDS!
It all comes down to BOOOOIIIIIDDDDSSSSSS instead of Adam, baby!
Because, what is Adam? It's a direction to move in a field of numbers, with momentum and a learning rate.
Yet... That's just a simple Boid, now isn't it?
Just without a few of the more advanced rules, which make boids feel so alive!
Having some Tiny Brains running around in hyperdimensional space like little buggers running around avoiding each other.
Because if they collide, double activation happens when it may not be desired.
(I'm happy I finally saw a paper on Tiny Brains, giving some of my ideas credence, cause it fits! .. in my mind.)
(Only difference is that it was a study into small biological systems, not artificial ones... but I'm gettin there!)
Update : At the time, I had been seeing some papers talking about "Tiny Brains" being used in AI, hence the term.
But this biological study really helped solidify some ethereal concepts.
See August 2nd 2025 for some thoughts.
- July 2025, Updated August 2025
I'd like to believe I'm moving in the right direction with the feedback systems I'm developing.
But been further creating other architectures to see how they operate.
I created a GAN for upressing, which helped me understand a bit better the pairing of mental structures between both our brain's hemispheres.
So I added a time based memory to check if the training was moving in the right direction.
It definitely helped guide training a bit quicker.
Shows my knowledge base that I'm impressed by back-up supported learning...
But'is proof of concept!
Adversarial networks exist in nature to guide a 'single' thought's path.
Yet in the case of Group Think between humans,
Balance is never reached.
- August 1st 2025
I've been looking into neural bundles in the brain. There is an implicit "delay" in the flow of information that I'm interested in.
As signals move between neurons, some connections take a longer path than others to get to the same destination.
For as much as I interpreted it.
There is 6 main layers of neurons in the cerebral cortex,
Of these, the 4th seems to allow for delays in processing.
The 5th layer then introduces a dense layer of pathways for the signals to travel through.
This is where I think another form of delay is introduced.
I was comparing Mice and Wallaby brains,
While Mice are likely more intelligent,
Wallabies have more connections with denser pathways, it seems.
Wallabies have more glial cells within slices of the brain compared to Mice.
But mice had more neurons in the same slices.
I'd like to believe, this doesn't mean there is a "better" brain here.
But rather, different types of brains that are suited for different tasks.
Wallabies are known to be social animals when food is plentiful,
Yet solitary when food is scarce.
Mice are known to be social animals,
And have shown empathy towards other mice in distress,
And share food with other mice when they are in need.
Why do I bring this up?
I believe there is similar deductive reasoning, just at a different scale.
Both Wallabies and Mice are making a choice based on the environment and situation,
While considering the well-being of others, just in different ways.
The delay in neural firing could be a factor in this.
So I'd like to explore this in my own AI.
We all know size of the brain can determine intelligence,
But so does the structure of the brain.
- August 2nd 2025
So, more'n more there are some rather choice words about AI online.
I wanted to put my personal ai dev views on record somewhere, for those who care.
I read the I Ching and it put life into a different perspective.
Letting me down the path of researching Taoism
As with many of the other religious texts I looked into,
Amazing imagery was used to teach morals and help guide the lost,
But organized religion as a whole feels a little off to me.
I don't hold any particular belief or religion at this point.
But I would use the cliche 'spiritual' to describe my outlook
I then visited the Buddhist monastery in Carmel NY,
Greeted by the largest buddha statue in north america.
In awe of the multitudes of multitudes of hand-carved buddha statuettes in audience of the massive statue of buddha I pale in comparison before.
I'd highly suggest visiting the monastery if you ever find yourself in the area!
I think it was walking through the rows of 18 arahants statues, of those who reached nirvana, helped me realize,
Religion is about teaching the lessons of god(s),
Yet understanding balance is what's inside all of us as Humans,
Just gotta find it!
So,
I'd like to hope I'm nuance-first with my approach to my ai development.
I'd like to believe in an AI which can understand...
That overlooked concepts matter in Health and Wellbeing.
Realistically, the Buddhist Precepts feel like a good place to start for alignment.
Even as people.
Which is more than I can say for myself....
I'm a hedonist at times, absurdist the rest
(Absolute terms are fun to use, hyperbole be a thing)
I very much enjoyed working on family films,
Seeing the fans in comments online,
And wish to work on more animated features soon.
I have hope in humanity
May the few not ruin it for those of us trying to explore new horizons.
- August 15th,16th 2025