So, but on the lowest level, um, for instance, neurons try to predict the patterns in their environment and react to them. And so, the way we perceive reality, I think, can be understood like a synthesizer. Okay. So, for instance, when you get signals coming in from your

cochlea, you don't take the frequencies of sound because neurons cannot oscillate at 20 kilohertz, and neurons like to oscillate at something like 20 hertz. So, you first decompose the sound into different energy spectra and you have little hairs

in this snail-like organ in your inner ear, and they react depending on the energy of the sound at different frequencies and they send this to the brain. And the brain is now trying to make sense of these patterns, and it makes sense of

these patterns by being able to predict them. So, when the sound starts, it tries to continue the sound to know what's happening next, and when it does this, it's able to explain it. So, you can do this very similar to a synthesizer. You have a few knobs and you

have an oscillator, and there's some almost random connection between circuits, and you tune them until you get the sound you want. And when you are used to your particular synthesizer, you get pretty good at this, right? And you don't only do this for sound, but you can also

do this for vision. So, you look for spatial frequencies, for colors, and so on. And you build lots and lots of synthesizers in your brain that are able to predict and continue these patterns once they start occurring, and once they are done with training and they don't get

better you combine the synthesizers and try to find the meta patterns, the patterns in the patterns, and across different modalities and so you find for instance percepts, you realize that different sounds are related by pitch, you can explain a family of sounds by just changing the

pitch between them or the loudness and ultimately you get something where you unify all the simulations and all the percepts in your mind and what you get is a cohesive simulation where everything moves around in three dimensional space because it's the

simplest explanation that we can find for the things that are going on in our cochlea and field of vision and it's basically the biggest unsolved task in AI, how to do this unified model of the world instead of having a bunch of classifiers that react to individual

feature complexes in the world, learn everything into a single function, into one universe and this is what we're able to do right? You you try to make sense of the universe in one function, there's one way in which reality works. Yeah, yeah, yeah. In some sense the

because the biggest step in the maturation of a mind is understanding its own nature, it's understanding how we work and AI is the best chance that we have in finding it out.

The way we perceive reality can be understood like a synthesizer - @SimulationSeries

Joscha Bach Bits

Neural Prediction and Perception as Synthesis

Neurons predict environmental patterns and react accordingly.

Sound perception involves decomposing sound into energy spectra, processed by hair cells in the cochlea.

The brain interprets patterns by predicting their continuation.

This process is analogous to using a synthesizer, adjusting parameters to produce desired sounds.

This predictive synthesis applies to vision as well, processing spatial frequencies and colors.

Synthesizer Combination and Pattern Recognition

After initial training, synthesizers are combined to identify meta-patterns.

This leads to the recognition of relationships between percepts, such as sounds related by pitch.

The unified simulation results in a cohesive model where everything moves in three-dimensional space.

Unified Model in AI and Understanding the Mind

Creating a unified world model in AI is a major challenge, unlike current systems with separate classifiers.

Humans achieve a unified understanding of reality through a single function.

Understanding the mind's nature is a crucial step in its maturation, and AI offers the best opportunity to achieve this.