“20% of neurons perform 80% of the brain's work”

Is our brain a marvel or just what it has been able to do?.

Oliver Thansan
Oliver Thansan
28 September 2023 Thursday 04:21
4 Reads
“20% of neurons perform 80% of the brain's work”

Is our brain a marvel or just what it has been able to do?

If I played you a Pink Floyd song and scanned your neural activity, I could design an algorithm with which to reproduce 90% of that activity...

Marvelous...

...So we could predict the effect of that music on his brain.

It would be seeing the music.

But it is no less interesting to observe that, if I play that same Pink Floyd music to a rat, we get the same graph.

That's not what I expected.

Because you, like all humans, have inherited the prejudices and terminology of two thousand years of philosophy, psychology, and concepts that neuroscience must now transcend to progress.

Soul, spirit, psyche, mind...?

All these concepts are useless and the scheme that the brain perceives reality and interprets it is also false...

It is not like this?

This scheme is invented by a human observer who observes brain activity and interprets it; but, in reality, the neurons in the brain do not observe anything or see anything or know anything: they are just particles that move by immediate impulses.

But don't we generate them?

Neural networks randomly and continuously generate and rehearse myriads of activity patterns that only when they benefit the organism gain content and do not disappear... The others, yes.

Does it send and receive continuous impulses?

It not only continuously sends orders to all the organs and muscles, but also forwards the responses it obtains and thus makes them interact with the environment. Only by moving the flower can we see it.

Would the brain – existing – be an incessant reality check?

But active, not passive. The brain is not a mere reception from the outside in, but rather it works from the inside out; Therefore, if we leave a baby at birth in its crib, motionless, without interacting with the environment, with a passive brain, it will never perceive anything.

Doesn't the brain perceive the world to interpret it later?

That scheme is wrong, even though it is two thousand years old...

Its alternative definition is complex.

Think of a baby babbling sounds: te-tut-ti-ted... And when you hear it, you confirm: Teddy! Yes, Teddy! Teddy! Teddy!

Like pa, pa, pee, pee, paap... Dad!

And so, the father confirms that babbling and gives it a permanent meaning for the baby. This is also how the brain acts: it tries round-trip circuits, first from the inside out, until it confirms one that works for it.

Until now, perception was described as the interpretation of signals...

...from the outside in, but it's not like that. There is no homunculus within the brain that is a passive perceiver of the signals from the senses and interprets them, but rather the neurons rehearse circuits until the action confirms their meaning and thus take root in it.

And so we are acquiring neural circuits to interpret reality?

This is how we interact with our body and senses and with the world, with signals back and forth: from the inside out and inside. The brain is a system in action, not a recipient of perceptions. Our discovery is that without action there is no perception.

And does this continuous trial-hit-error of the brain enable it to adapt?

It is like, by superimposing billions of circuits as we live, we adapt by anticipating the changing circumstances of the environment. But don't substitute evolution for God. The brain is not going anywhere: evolution is not progressing towards any goal. It only adapts to survive.

Aren't we wiser than a monkey?

The brain is a medium and long-term prediction machine; but the bee's brain is also efficient for the honeycomb and correctly predicts it, but only in the very short term, because it does not need more.

Do all neurons participate in all neural networks?

On the contrary, let's say that 20% of neurons carry out 80% of brain activity, which is very hierarchical.

But neurons are very diverse.

And that diversity makes our neural network powerful, stable and resilient; but not egalitarian, but hierarchical. That's why engineers make mistakes when they design artificial intelligence networks with equal members.

There are two ways of thinking, says Kahneman: think fast or think slowly?

There are two ways of thinking, yes. And the first impression that the fast one gives us is not the best: the student who seems weak on the first day can be brilliant after five years. Thinking quickly serves motor reactions, but good judgment requires time.