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!
For my 2026 goal, I've been exploring Graph Attention Network (GAT) artificial intelligence.
As GATs allow me to treat 'concepts' as 'objects', rather than sections of words/pixels as a tensor or 'piece of a chunk of a concept'.
GATs are a type of neural network that considers the relationships between data points.
As a type of Graph Neural Network (GNN),
Its best for predicting connections between ideas / things / data in a system.
I've been working on a general-purpose neuron that adjusts its own connections during prediction;
So the same system could learn my voice on the fly, as well as sensor signals connected to the Jetson computer.
Since its the Structure in a GAT that causes regions of neural activation based on stimuli,
It forms a result (prediction) after subsequent activations, as-though compounding ripples in a pond.
Rather than a field of numbers aligning to yield a prediction,
It's the structure of neural connections which manipulates the data.
I've been going in a direction that should yield a similar result to a Recurrent Neural Network (RNN), but with a different mental structure.
With that general-purpose neuron, I can provide text, images, audio histograms, etc. to the network.
RNNs can be used for/in many types of ai,
Best for detecting patterns in sequential data,
Like time-based events or words in text. They are the basis for many types of ai, like LSTMs.
The GAT will create connections from initial random data points, sample the differences, then pass the 'prediction' forward and 'back' in the chain, and adjust the connections based on their revisit to the same data in the current 'prediction'.
Relying on localized regions of sub-networks to recurrently process the data
It should be self-taught discrimination of attention between neurons;
Like in the human brain.
Please note, I haven't mentioned the transformer for this GAT.
It was byte-pair 'tensors' encoded text block that I fed into the GAT.
The GAT then found connections between the occurance of 'tensors' in "sessions" of other 'tensors'.
So the nodes are "occurences" of neighboring 'tensors' in the text block.
It was like a, "find my neighbors" type of search for the GAT.
What's not visualized here?
The "ripple" through the GAT nodes during training epochs.
... And the attributes of the nodes.
I have a new GAT use-case in mind that should better show how "language connects" in a visual way soon.
What it's trying to do?
Link multiple nodes together in series to recreate the "rule" for those tensor neighbors.
By recreating 'use cases' of the 'tensors' in the text block.
Which is why there are so few nodes here.
What I didn't know? This is more of an MPNN than a GAT.
A message passing neural network (MPNN) is a type of GNN that passes messages between nodes to update their states.
It could really use some better visuals for this anyhow....
But hey!
I'm a toys'r'us kid after all, so....
Hooked on Phonics worked for me!
GATs are a type of neural network that considers the relationships between data points.
As a type of Graph Neural Network (GNN),
Its best for predicting connections between ideas / things / data in a system.
GNNs are commonly used for "Recommendation Systems",
Hey, you might know Jim Bob McGee!!
But GATs could be used for so much more!
Hey, you might know Jim Bob McGee!!
But GATs could be used for so much more!
I've been working on a general-purpose neuron that adjusts its own connections during prediction;
So the same system could learn my voice on the fly, as well as sensor signals connected to the Jetson computer.
Since its the Structure in a GAT that causes regions of neural activation based on stimuli,
It forms a result (prediction) after subsequent activations, as-though compounding ripples in a pond.
Rather than a field of numbers aligning to yield a prediction,
It's the structure of neural connections which manipulates the data.
I've been going in a direction that should yield a similar result to a Recurrent Neural Network (RNN), but with a different mental structure.
With that general-purpose neuron, I can provide text, images, audio histograms, etc. to the network.
RNNs can be used for/in many types of ai,
Best for detecting patterns in sequential data,
Like time-based events or words in text. They are the basis for many types of ai, like LSTMs.
The GAT will create connections from initial random data points, sample the differences, then pass the 'prediction' forward and 'back' in the chain, and adjust the connections based on their revisit to the same data in the current 'prediction'.
Relying on localized regions of sub-networks to recurrently process the data
It should be self-taught discrimination of attention between neurons;
Like in the human brain.
(When the purple circles go red in the GAT video, first vid)
Please note, I haven't mentioned the transformer for this GAT.
It was byte-pair 'tensors' encoded text block that I fed into the GAT.
The GAT then found connections between the occurance of 'tensors' in "sessions" of other 'tensors'.
So the nodes are "occurences" of neighboring 'tensors' in the text block.
It was like a, "find my neighbors" type of search for the GAT.
What's not visualized here?
The "ripple" through the GAT nodes during training epochs.
... And the attributes of the nodes.
I have a new GAT use-case in mind that should better show how "language connects" in a visual way soon.
What it's trying to do?
Link multiple nodes together in series to recreate the "rule" for those tensor neighbors.
By recreating 'use cases' of the 'tensors' in the text block.
Which is why there are so few nodes here.
What I didn't know? This is more of an MPNN than a GAT.
A message passing neural network (MPNN) is a type of GNN that passes messages between nodes to update their states.
It could really use some better visuals for this anyhow....
But hey!
I'm a toys'r'us kid after all, so....
Hooked on Phonics worked for me!