Sleep itself is characterized by coherent neural activity— the large number of brain regions with synchronized neural activity. The slow waves where huge numbers are all firing close together in a rhythm. Low frequency and high amplitude delta brainwaves (1-2 hertz).

Complex adaptive brain activity requires more complex firing than a simple rhythmic frequency. So, in a way, the complex activity must be stopped in order to support global synchrony.

Why would our neurons want to all fire synchronously? Well, it is healthy for neurons to fire together in a causal manner— neurons release growth hormones then. That neural growth during synchronized firing is the basis of “neurons that fire together wire together.” And it doesn’t seem coincidental that a successfully predicted model feels good, as in the case of successfully throw a ball in a basket. Neurons are trying to predict other neuron firing and respond to it. If they are unable to effectively, they may become like the 1/3 of our baby neurons in the cortex — they will be pruned and die.

Good feelings is positive reinforcement—behaviors leading to good feelings get reinforcement. The feeling of harmony or harmonization, where we have to balance a broad set of internal neural impulses, feels good when we do it well. We feel harmony in music — and in our own internal sensory resonance to the world.

Hypothesis 1: the harmonization of neural activity might cause conscious feelings due to the convergence of the activity to platonic forms (see Platonic Representation Hypothesis in LLM research).

Returning to sleep — this is a proposal for why sleep feels good. Synchronization might intrinsically feel good. But because the sleep also disrupts your working memory contextual attunements (ie, whatever your day was about) - taking your brain into deep synchrony — it strengthens the overall dendritic connections between the synchronizing neurons.

And, because it wears off the edges of your previous experiences — you can return refreshed.

In this way, sleep seems to contribute to the overall integrity of the operation of our intelligence. Without it, we lose integrity and internal harmony.

And yet, not sleeping is one of my favorite drugs. Can be a major performance enhancer, even if it is variable.

Hypothesis 2: Not sleeping increases the (statistical) temperature of the brain.

It's also a fundamental misunderstanding of how LLMs work, mixing up inference with training.

Descartes compared the human mind to waterworks and hydraulic machines, other authors used mechanical clocks, telegraph systems, digital computers, and (in the recent decades) neural networks.

In the end it's all computing and to a degree all of those models serve as good analogies to the wetware, one just needs to avoid drawing wild conclusions from it.

I'm sure there will be new analogies in the future as our tech progresses.

We don't literally train on today's prompts while we sleep, but there actually _are_ some _computing_ tasks going on in our brains at that time that seem to be important for the system.

    private static readonly final sleep()

By the way, even Cnidaria (jellyfish etc) exhibit sleep-wake cycles [1]. They don't have a brain, but they do have a nervous system. Maybe the first animal with nervous system (a common ancestor of Cnidaria and Bilateria) was the first to have a sleep-wake cycle.

I don't understand the current research on mitochondria, but it sounds as if sleep has to do with how neurons work.

1: https://link.springer.com/chapter/10.1007/978-3-031-62723-1_...

Almost the same here but it's not a deep nap for me. I relax, start seeing dream-like images in my mind (yet still drifting into-out of conscious awareness), then in ~15 minutes I feel energy build up and am ready to jump up and go.

I would say that the darn alarm clock prevented me from completing a sleep cycle properly in the morning, and now I did complete it and made my brain happy.

Or maybe going all the way on and mostly-off with your mitochondria, even specifically with your brain mitochondria, really is that much more efficient than having half of them offline (but still consuming energy for upkeep) at any time. The brain is a big ol energy hog, after all.

The brain has uniquely high specific power requirements per gram of dry weight. Not even the heart is this power-hungry. This surely places a lot of uniquely high metabolic stress on the neural cells.

And neural cells are long-living, so they can't be easily replaced throughout the lifetime. So their housekeeping has to be very thorough, carefully cleaning up all the waste products.

So this hypothesis actually makes a lot of sense.

It's no more nuts than being awake given how much energy vigilance costs.

Maybe alcoholics do enjoy drinking. But working 60 hours a week isn't going to cause brain damage and liver failure on its own. Productivity isn't a chemical or vice.

Being unconscious, or in a coma, in a hospital bed is more akin to being asleep, which is why you can be in a coma for years without dying.

So eg https://translational-medicine.biomedcentral.com/articles/10... reviewed 19 studies, many of which did find "evidence of mitochondria problems", but concluded:

> ...it is difficult to establish the role of mitochondria in the pathomechanisms of ME/CFS/SEID due to inconsistencies across the studies. Future well-designed studies using the same ME/CFS/SEID diagnostic criteria and analysis methods are required to determine possible mitochondrial involvement in the pathomechanisms of ME/CFS/SEID. [...] There is consistent genomic research suggesting that ME/CFS/SEID is not a primary mitochondrial disorder, however, mitochondrial decline might occur due to secondary effects of other disrupted pathways. [...] As population samples were small, these results should be interpreted cautiously.

I wouldn't summarise that as "no evidence". It's more like "ME/CFS doesn't seem to be a genetic disorder causing defective mitochondria, and the mitochondria look the same, but they seem to function differently for some reason even if we lack enough data to figure out why yet". Note that, eg, of the 19 studies reviews, 5 tried to check for differences in mitochondrial respiration between ME/CFS patients and healthy controls, and 4 of the 5 found notable differences; one study was able to reliably detect if a cell sample came from a ME/CFS patient or a healthy control based on measuring mitochondrial respiration.

I don't know that's enough to fully reject the null hypothesis just yet, but it's certainly not clear we can accept it either.