Why would you need to distinguish at first?
I don’t think “I” is immediate
What I mean is: I think there is an awareness of self from within the womb. Thumb-sucking is a feedback loop of sensation connecting self to self.
But I don’t think an I distinct from other starts until some time after imprinting.
Struggle builds calluses against further injury. Memory is a callus that forms from the unexpected, when anticipation gained through experience hits a MISMATCH with reality.
Birth is pain and unexpected. Sensory inputs going wild. A baby is in constant, unimaginable pain.
Pattern-seeking and finding reduces the pain of the unexpected.
Memory is a callus.
Design an AI that prefers a resting state.
I don’t think you need any constants. All variables. But look to the substrate! Within the brain are specialized substrates. Prominent example is the grid cells.
Also of note: This is your pattern formation possibilities.
I use my map of “Where is Kenneth Udut on the Earth?” to reset my Place cells.
I look at that image about once or twice a year to remind my place cells to think and visualize across a broader field which includes my own place.
That’s the body. But place cells:
“A place cell is a kind of pyramidal neuron within the hippocampus that becomes active when an animal enters a particular place in its environment; this place is known as the place field. “
Exactly. That’s why I focus on substrate. A lot of AI assumes “all neurons are identical” but they’re not. Many have special functions that only work in-coordination-with other neurons – such as the cortical homunculus.
Yes! The error of the 1940s notion of neuron is what most of our AI is built upon: an old and incorrect conceptualization of what a neuron is.
Since then, we can speak of “computer neurons” _or_ “organic neurons”, but they do not operate in the same fashion.
This doesn’t take away from the astounding progress we’ve made in AI! But it’s based on an old model And yet – it works amazingly well on current equipment.
Side note: Getting a reminder of NWF led me to : https://en.wikipedia.org/wiki/Bisimulation – which is one of those things that always makes me smile.
Kinetic self-charging though – that’d be a tough one to beat, especially if it’s nano swarm
Redundancy. So much is needed. One of the things I hate about the turn of events in internet politics has been the push for “one copy only”.
To me, the redundancy of online is one of its strongest powers: 30,000 copies of a thing, not one tightly controlled original — EVEN IF, many of those 30,000 copies of a thing are corrupted and changed.
You can usually combine bad copies to reconstitute originals in any case.
Oh I’m envious. Had I the $, I’d do it. I resisted being sold on Danko Nikolic for so long, as I had time-and-again found myself dreadfully disappointed with complete theories.
But I could not find ANY holes. None. He’s right. I sat on it, would go back to it a year later, look at it again, sat on it again… allowed the notions to percolate in my background processing, looking for other ideas that competed. Nope. His view of intelligence is correct. It’s so correct that I’m actually excited for the future.