This framework wasn't born from research and reading. It was forged from trying to understand.

To understand why people keep hurting each other, even when they are trying to do better. Even when they hate hurting each other. Even when the solution should be easy, something like just replying to a message.

Brain neural network visualization

Fig. 1 — Neural network optimisation landscape

I could not make it make sense. No matter what I did. Until one day I realised there is an energy limit. When we hit it everything else goes out the window.

Then I saw how we are all trying to do better. We are trying not to hit our energy limit.

Then it hit me, gradient descent, optimisation theory. As soon as I saw that I saw the landscape. Then I knew the maths. I'd always intuitively understood human behaviour well, it was one of my obsessions. But suddenly I could map that intuition onto maths. It was a breakthrough. I knew I was onto something.

My only real problem at the beginning was I couldn't work out if the y axis was energy or "happiness". It kept flipping between them in my mind. I sorted that one out. It was both. Well kind of.

And then how to explain personality based motivations. I sorted that one out too. Though it's not for this theory right now. I will write that one up later.

As I unfolded the theory it started aligning with existing ideas—attachment theory, trauma responses, controlling behaviour. But it didn't just match them. It unified them. Gave them a shared structure.

So this is my theory. Mapped out through explaining human beings and social dynamics. Which is how it was developed. But I also have the maths. And this page is about the deeper intuitions. The intuitions about why the brain is optimising an emotion based landscape within a very strict energy limit.

Here are my postulates.

1
Our brains work as a neural network. Rewards are given for actions that make our survival more likely. Adverse states are those that make survival less likely. This functions as somewhat of a cost function that is minimised over our lifetime.
2
Emotions evolved as a proxy for survival. Brain chemicals give neurons critical information on the current state of the "cost function".
3
Our emotional reactions are largely innate, based on evolutionary history.
4
Brain chemicals allow the reward to move to a continuous function, away from a binary survive or die.
5
Each brain chemical have different properties, and they interact with each other. Our emotional state is our internal representation of those brain chemicals.
6
Signals process through the brain, neuron to neuron. This process results in actions, thoughts and learning. Learning is online.
7
This cost function is local to each neuron. It only sees it's only states. Through it's only inputs and outputs.
8
There is no clear distinction between signal propagation and final actions. If neurons that facilitation actions are triggered that action will take place.
9
Brain regions evolved through the process of branching pathways adding higher order cognition. Over time if these branches gave better outcomes, they were rewarded and developed into what we see as distinct brain regions.
10
Thinking about future (or past actions) simulates their emotional experience. This allows us to make decision on future states. Make present moment decisions based on longer term goals.
11
We can consciously choose what we think (to an extent) about which shapes the current landscape, therefore shifting out actions. However our actions are based on an emotional landscape not interlectual.
12
We build sophisticated "models" of our emotional experiences (landscape). This happens mostly within our higher order cognition regions. This allows us to make better decisions in both the moment, and in our future projections.
13
We have an instantaneous processing limit. Above this higher order cognition regions (such as PFC) go off line.
14
If an experience is above this limit we are not able to process and integrate it sufficiently. In addition the next time a signal travels down these pathways it is likely they will release the same chemicals (e.g. cortisol) causing us to once again hit the energy limit. Reinforcing that pathways, actions and experiences that trigger this are dangerous and should be avoided.
15
High cortisol shuts down higher order cognition. These high danger triggers are encoded broadly and deeply. This creates a low precision representation, which keeps us away from broadly similar dangerous situations in the future.
16
The brain is rarely disordered. The learning and actions it takes generally make sense based on past experiences and current signals.
17
Logical and emotional pathways are integrated. Logical pathways link emotional concepts and add new meaning and understanding. This is due to Hebbian learning between distinct regions.
18
The brain just is. It doesn't calculate gradients or derivatives anymore than the universe solves Schrödinger's equation.

As a person I'm always thinking, hypothesising. My mind is full of intuitions and ideas. I've always struggled with the gap between what I'm seeing and what I can prove to others.

I always thought if I was more logical, had more evidence, explained more clearly, I'd be able to get others on board.

My theory is my success. I finally found a way to communicate what I feel is true. But it's also helped me see what was really going on. I was always seeking. I believed if others saw things my way they would be nicer, they would stop making the same mistakes, the world would slowly become kinder one person at a time. My theory is the accumulation of this. But now it's the new thing I have to find a way to communicate.

I'm a scientist. I want to progress science. I want to take this theory and write papers and get it into the world. But I also know that's a life's work. And it's not my life. My true satisfaction comes from helping people make better choices. While also knowing they probably won't, well not in obvious ways. That's my north star. Not chasing the permission to see the world the way I do.

This website is for my ideas. Wherever they may lead.

The brain is rarely disordered. The learning and actions it takes generally make sense based on past experiences and current signals.