Translation to machine language inside the mind

Translation to “machine language” inside the human brain is really interesting when you think about how you are interacting with other people in the world.

I would argue that my social skills are good.

To explain why is a fairly complicated process.


I think that I’m good at understanding the underlying meaning of language when someone speech. Meaning I both understand the actual words but more importantly I understand the motivation behind the words. So when people speak with me the communication is fairly lossless. I try my best to absorb all the data.

The next step is translating that data into something meaningful to act on.

This is incredibly complex and is related to theory of mind, the idea that inside your mind you have a mental representation of another person, as well as default models of behaviour. You need that representation of the other person to make sense of the words they speak, because you need to understand how their words relate to your view of the world.

You then need to translate their words using your mental representation of them into your own machine language, the language that your mind speaks.

Once that is done the computations are run inside your mind, their language is processed and you are working on a response.

To make a good response is also incredibly complicated because your response should both be accurate, kind and sincere and take into account the other persons knowledge of the world (not your own) to form a coherent response.

Then you need to speech in an articulate manner with enough information to keep the other person interested while remembering to keep all your atoms and molecules working.

Pretty hard stuff.

Overfitting biological neural networks

I think it could be good if we spent a minute talking about overfitting.
Because overfitting is the biggest problems with artificial neural networks it is likely true that they are also a huge problem in biological neural networks (your brain)
Your brain often works with the map of an object instead of the territory. This means that the bran often spends much more time thinking about the state of the world than acting in it.
This can lead to interesting errors. Because we are seldom checking our models in the real world.
To quote some math
“The error on the training set (your inner representation of the world) is driven to a very small value, but when new data is presented to the network the error is large. The network has memorized the training examples, but it has not learned to generalize to new situations (act in the real world).
Here are some example:
You meet a new person, who you think is fantastic in every way. After the meeting you reflect about how amazing and wonderful they are, they seem to be perfect. When you meet again it turns out they are just a normal person (aren’t we all?).
You meet a new person, who you think is terrible. After the meeting you reflect about how awful they are, they seem to be dreadful. When you meet again it turns out they are just a normal person (aren’t we all?).
You are sure that people in the category group X has attribute Y that makes them better than people in the category group Z who has attribute Q.
Of course I make the same mistakes too. Because you can never get rid of your own bias. All you can hope for is to become more aware of it. If you believe that you don’t have any bias you are wrong. Plain and simple.

Neuronal soup

Your short term abilities as a person (your thinking, actions, skills, emotions, fears and loves and your entire view of reality) is dependent on the content of the neuronal soup that is inside your head.
This complex soup holds everything inside your mind during a fairly short time period and is a combination of the unique makeup of oyxgenation levels, blood flow, neurotransmitters and blood glucose levels (and more stuff).
It is likely true that the unique makeup of neurotransmitters is what encodes as memories. These memories keeps a time series going in the brain with each moment creating a distinct marking. You can think of this timeserie as an infinite series of snapshot that encodes reality at an incredibly high framerate.
You must understand that everything you ever experience gets encoded. So it might be useful to spend a moment thinking about what you experience.
Over time this soup fasten like the baking of a bread and our day to day experiences form memories that hopefully helps us navigate the world better. This memories improves our ability to reason and think about the world once they becomes heuristic tools for thinking.
This concept can help explain two types of intelligence.
Fluid intelligence, the quick wit that is often accompanied with youth (your soup is excellent today!) and crystallised intelligence, the wisdom that comes with age as our knowledge expend (if I know more it is because I have stood on the shoulder of giants).
So how can we use this knowledge to learn better today? The solution is obvious.
1. Make sure your neuronal soup is as great as it can be. Surround yourself with people who are loving and help you learn. Expand your mind by trying out things that are at the edge of your ability. Take care of your physical body through exercise and appropriate rest. Hydrate, eat and sleep well.
2. Protect your neuronal soup from the tyranny of the world. Abstain from acquiring more knowledge about the madness of the world if the knowledge isn’t absolutely necessary.
3. Understand how you can get your mind back to a place of good learning. For me it’s a combination of love and courage. Love because anyone who loves you stores a copy of you inside their mind that is useful for remembering the greatness that is you. Courage because a mind without fear can learn anything.

Neural pruning

Had a fascination conversation about neural pruning with my roomie.
I like the idea of explaining ideas of mental illnesses through this lens. Neural pruning is your brains ability to remove connections that are no longer useful. To prune ineffective neural connections.
A metaphor that is familiar (but unproven scientifically) would be heartache. Your mind is pruning connections to someone that is no longer present in your life. It hurts.
Problems with neural pruning was linked to schizophrenia but you can view it as a framework for looking at different aspects of mental health.
The basic idea would be the following.
Making mistakes when we think doesn’t matter all that much. We all make mistakes when we think. It is a part of learning. The problems arise when we stop learning from our mistakes. When we stop pruning our neurons based on feedback from our environment.
At the same time, it’s a complicated tightrope. If you prune too many neurons you might let go of your dreams. Because dreams are the things we believe about the future without proof. If you prune too little you suffer delusion.
However you could argue that entire societies run on delusion/dreams. The american dream, The Swedish welfare state.
Many people prune their dreams away and live stable low risk lives. Others don’t and increase potential risk but also potential reward.
(For those of you have been paying attention you might think that I’m writing about neuroscience from many different viewpoints. This is because neuroscience is an incredibly complex field and you need to view it from different lenses to get a grasp of the problem).
Here a link to the article.

Predicting the future

The minds goal is to try its best to predict the future.

It does this by creating a mental model of the world and everything in it including you. This mental model of you is likely the most important construct inside your mind because every computation inside your mind goes through this construct.

Every action you take is determined by your own belief in who you are.

We can simplify this further by stating the following.

Your actions in the world are a determined by a collection of I statements about yourself.

I am smart, I am funny, I am kind.

We can translate these into math.

W1 = I am smart = 22

W2 = I am funny = 20

W3 = I am kind  = 18

You = W1 + W2 + W3

In a mathematical neural network the computations of the networks are determined by the values of weighs and biases within the network.

So when we compute you we take everything you think you are and we compute the output of that to get a value. So you are 22 + 20 + 18 = 60.

Negative values attached to the self are catastrophic for mental health because every time you have to think you will experience negative feelings because you will get negative output when you compute yourself.

I’m pretty certain that just reading negative stimuli slightly changes the weights and biases within your mind but perhaps some people aren’t sensitive enough to pick up on this slight change in the computational models within the mind.

If we would add “Im a fraud” to the mix, you will suddenly have a new computation

W4 = I’m a fraud = – 25

You = W1 + W2 + W3 + W4

You = 35

To further complicate things you can decide on a meaning for a word in your mind that makes it so that it doesn’t influence you negatively. A famous example.

“Here’s to the crazy ones. The misfits. The rebels. The troublemakers. The round pegs in the square holes.

Here being crazy is good.

No matter what you perceive yourself to be your belief in yourself influence you.

These beliefs in yourself determines what actions you perform and what actions you don’t perform. They are a sort of rule based schematic of how to navigate life that you have imposed on yourself to handle the absurd complexity that is life.

Not having an attribute for yourself will also influence your actions. This is an unknown unknown. I’ve painted for almost a year and I self identify as an artist. This highly influence my thought process and I’ve had a lot of thoughts in the last year that I wouldn’t have had if I didn’t identify as an artist.

So how do we solve this problem? You can’t really lie to yourself too much

because your mind will require proof that you are doing something. If you say that you are a runner you need to run. But you can start by deciding to be a runner and then start running. There’s a gap here that you can use to get started with any activity.

Once you self identify as something and determine that it is important for you to be that thing your mind will give you the resources you need to pursue that goal. However nothing comes easy. I spent months doing nothing but paint to learn how to paint. This is how it works regardless of what you need to learn.

But it is important to understand that you have the choice to be whoever you want to be. You can literally change overnight and decide to be a concert pianist and after a few thousands hours of practice you will probably be pretty good.

By convincing your mind that you are something you will gain an edge and over time you will become what you convinced yourself that you are.

As the old saying goes.

Fake it till you make it.