Rather Suspect

There is our internal world and the external world. But why should this be?

Or to say it another way, how (and why) would self-awareness arise?

From an evolution perspective, perhaps it might have gone like this. A set of nerve sells begin to sense the external world, and these aid in finding better food and evading more predators. And then — perhaps — over time a few of these cells accidentally mutate, and instead of sensing the outside world, they sense the other nerve cells. Nothing intentional here (as with all of evolution), a cell just gets mis-wired and connects to another cell rather than the outside world.

So now, what? The first cell smells food. The second cell senses the first cell smelling food.

A third cell mutates yet again and this one senses the second cell sensing the first cell smelling food. Tangled wiring is mathematically more likely than neat, ordered wires, and so of course the axons and dendrites end up all intertwined.

I suppose from here the key is, this tangling must lead to some evolutional advantage. And it likely does: to use a physics example, the first cell feels pressure but the second or third will track the acceleration or deceleration of that pressure. So now we have two data points about a food source rather than just one, and can better distinguish between nourishment and poison.

What’s interesting in this model is the concepts of “time” and “self” emerge again, the two components of consciousness. Notably, neither are hard concepts in the external world, but simply emergent properties of this organically-grown sensing machine. Time arises only because the information travels in a serial path from first cell to second to third and back to second; if the second cell had positioned itself properly, on the surface of the organism, there would be no need for this serialization. And the “self” then arises when the information goes around the loop a second time.

In this model, neither Time nor Self exist in the external environment, but only because of these loops.

Today we create so-called AIs by building neural networks filled with feedback loops. Makes sense, but two questions arise from this. First, what is needed for those systems to create those properties of Time and Self? And second, what other wiring geometries could lead to an evolutionary advantage that don’t involve loops? The second would enable “thinking” that is both timeless and selfless.