Brain Science Research: What Bayesian Inference is NOT for Mental Health | VR
The brain never makes predictions. If it did, predictions would be responsible for conditions like autoimmune diseases, because rather than the brain’s hypothalamic–pituitary–adrenal (HPA) axis [said to be involved with the immune system] having precision, it would err by predicting that immune cells can attack nerves or body’s own cell, because they are similar to foreign agents, as a mistake of faulty prior beliefs.
Instead, autoimmune diseases are similar to forgetting, such that the way control or regulation [as memory] should be made for limits, extents or precision, isn’t remembered in the moment, leading to the self-attack.
The brain does not have any internal model of the world that was not picked up through sensory relay pathway. There is no external model that does not go through that pathway.
There is nothing like prediction error in the brain, correcting or updating differences. These labels apply fancifully for certain purposes, but there is no functional basis of thought and memory, that aligns with those, constantly, across the brain.
There is no psychiatric disorder where prediction error can be used to accurately model what is wrong, because prediction does not mention the quantity of relay, nor does it discuss the construct [thought and memory] of cells and molecules, for how they decide the stretches of mental illnesses, psychiatric disorders or for mental health.
There are several experiences individuals anticipate or expect things. In the memory, it works like whatever is incoming, splits in two and one goes ahead, like it had previously, before the rest follows, along the same path or differently, all happening so fast. It does around what was already picked up and stored in memory locations that act along inputs.
It is not all experiences that the split happens. When it does, expectation or anticipation follows, when it doesn’t, nothing. The brain is not constantly generating predictions, as said everywhere. The brain is not a prediction machine, hallucination machine, making best guesses or what the free energy principle describes.
The free energy principle is based on Bayesian inference, sometimes related to the Bayesian brain, about prior belief, or internal and external model, but they are weak when applied to mental health.
Anticipation, as expected or not, could be responsible for delight or disappointment. There are anticipations of things going wrong that do and causes problems.
There are those that don’t that causes delight. Disappointment or delight could harm or benefit mental health. There is a split of stores in memory going-before responsible, functionally, to expect.
It is a function of memory, storing thought, working within their rules, but not predictions or guesses.
What is described as “the integration of feedforward (sensory) and feedback (top-down) neuronal signals is a principal function of the neocortex” is more directly the process of how sensory inputs become thought, and how thought becomes what gets relayed for interpretation, which is knowing feeling and reaction. In the brain, it is no longer signals, senses become a new identity which gets stored in memory, and useful in interactions with the world.
The brain is not an organ of labels, but an organ of constants, with overalls, by cells and molecules constructed for thought and memory, holding promise.