Thank you Dr. Nikolic. I’ll look at it, quite gladly. As a seeker of “what’s correct” in models of reality, I’m cautious because a model can be “almost correct” but have a fatal flaw in reasoning, or at least something that raises a “note of caution”. That doesn’t mean I won’t advocate it but I want to be aware.
Yet, with Practopoiesis, I find no such flaw.
The closest model that “speaks to me” in the realm of cognitive development that is “almost perfect” to me is “Embodied Cognition” by George Lakoff along with similar ideas by others that followed him. Yet, I am aware of some criticisms of the “family of” embodied cognition concepts and I keep those in mind, even while unhesitatingly endorsing it.
Yet your Practopoiesis has no flaws that I can find.
From a common sense perspective, any intelligence should necessarily have available everything it has learned thus far as a tool in ongoing salience and decision-making and be capable of revising any portion of its intelligence should feedback (internal or external, it does not matter for it is one system) show that greater affordance is given upon revision.
Affordance awareness brings agility to novelty.
Your model includes that “bringing everything” feature. Evolution, Genetics, individual cognitive development through the lifespan are all in some way modifiable in Practopoiesis. Evolution may be a slow facet of intelligence, genetic expression a faster facet of intelligence, and moment-to-moment neural activity the fastest facet of intelligence, but all work together systemically with and through each other as continual feedback is processed from the system which includes its environment..
This, at least, is how I’m understanding Practopoiesis and how I’ve been explaining it to others, and I’ll keep studying just in case I get something wrong.
All in all, I want to say thank you.
I will read the PDF at http://www.danko-nikolic.com/anapoiesis/ to refresh my memory. I read your materials first in 2014 then 2015 (“AI Kindergarten” was making the rounds on Youtube and in viral shares) and remember finding it agreeable then, but as I had “put it aside” to let my unconscious “work on it” in the background, I’ve forgotten details.Anapoiesis is the “verticalness” of the system, yes? (loose/flexible models are stored in long term memory and are instantly available to be configured specifically in the working memory as needs and affordances dictate)?
That’s how I remember it but I’m going to read your PDF again now.
Mainstream neuroscience seems generally ‘stuck’ in segmented computational models. What they miss is that computers were based upon our understanding of human cognition at the time. Human modeled computers after how they thought the brain functioned.
Yet that’s been forgotten and “Your brain is like a computer” dominates.
But really it’s “A computer is like your (old model of your) brain”. The greater form can’t be modeled from the lesser form.
However, if AI is built with anapoeisis as part of the cognitive model instead of outdated computational brain models (the computer that was modeled after the brain becomes the model FOR the brain – all quite absurdly circular) – AI will reap the missing benefits that ACTUAL intelligence brings that current computers are missing.
I looked Robots Go Mental just last week and I think you got it right. Intuitively “seems right” and what I’ve read so far seems right. At the very least, it aligns with my deep models of “what seems right”.
[I was a fan of Connectionism since first learning of it in 1990, the untapped potential of truly Parallel processing, the agility of Markov chains, and have enjoyed watching growth in AI since around 2002 again, after the long “winter of AI” in the 90s. Yet persistent AI models are too optimistic (“complexity itself magically brings intelligence”) or too tied to unnecessarily modular models (computer-based metaphors)]
Apologies for writing so much but as you can see, I’m clearly a fan of your work – and that’s why I’m cautious But nothing you’ve presented that I’ve seen has given me reason to doubt nor have I found any criticism of your work thus far.
Thank you for your time and I’ll revisit Anapoiesis as well as read through the peer reviewed form of Practopoiesis.