Hi! We are Raffi and Marley, and we are not investors in neurotechnology.

The substrate

Your brain is a (very) large network of neurons. It is sparsely connected. Bidirectionally coupled neurons (2-cycles) are overrepresented by several magnitudes compared to a random network.

the synaptic connectivity matrix has the following properties: it is sparse, with a large fraction of zero synaptic weights ('potential' synapses); bidirectionally coupled pairs of neurons are over-represented in comparison to a random network Whereas the human eye takes in about one gigabit of raw visual information every second, we end up using only a few tens of bits to guide our behavior

Mice have more synapses per neuron? linear in synapses, It's been discovered that primates actually have less connections per neuron than mice Stateful

There is a deterministic mapping from genes to neurons in C. Elegans, a creature with 302 neurons. In humans, the network is dynamic, lacks predetermination. Notice that there is more information in your 10^11 neurons than your 10^9-bp genome. Nerves grow through a random search strategy, which you can get an intuition for by looking at the vessels of your arms.

Your own brain structure will look different from your brain structure at t-100ms, to say nothing of the differences from one person to another.

[highly complex solution to a problem, if you "solve" one brain you don't have the key to another]

In fact, the brain is stateful in two ways. In the physical structure of its connections, and in the "on" or "off" state of its neurons. The program and the variables.

Funny image of your variables modifying your code, because You might guess that the input and output neurons are more interpretable

When you repeat a phone number in your head, the memory is represented in neuron firing that lasts just as long as you do it. When you recognize the word "salubrious" after going years without, you owe it to a physical network somewhere in your brain.

(fast, dynamic)

You can fall asleep and still know English in the morning

Notice that we don't add or take away neurons to represent state. Neurogenesis happens once in childhood, then stops in adulthood.* Except, for ~unknown reasons~, in the hippocampus and olfactory bulb

Big Data regime -- It is obvious what is hard.

Clearly that iinformation How does your brain work?

What have we figured out, what path did humans take to our current understanding?

How many neurons do we need to record to create a reasonable interface? Why can't we do everything noninvasively? What are the physical constraints for neural imaging?

What is EEG? What is it generated by?

nearly a century after the neuroanatomical organization of the cortex was first defined, its basic logic remains unknown The atoms of neural computation

in one view of cortical microcircuitry, pre-structured connectivity will provide “inductive biases” for learning and inference computations, and in that view, we may expect different structured inductive biases in different areas (which process data with different properties and towards different computational goals) to be reflected in differing local microcircuit connectomes across areas