David Gelernter – now-legendary computer scientist and AI authority
Read the first 25 or so pages, because it was listed here.
Not particularly impressed.
Seems to have made a large number of errors in those first few pages, so not going to go any further unless someone else comes up with a good reason to.
In the introduction he states that Turing suggested – “rational thinking was the manifestation of intelligence”. I strongly suspect Turing was wrong on this. I suspect that intelligence is in reality the ability to find and model pattern that exists in reality. Rationality is one very powerful tool in doing that, and it is only one tool. There are other heuristic and pattern based approaches that are far more powerful in certain contexts, that we all use.
The more tools one has in the toolbox the better, as far as I am concerned.
Reality seems to be sufficiently complex that all such models must be based on sets of heuristics (best guesses, useful rules of thumb), rather than any sort of absolute first principles.
Reality seems to be based upon stochastic systems at its lowest orders, but at higher orders those probability distributions deliver something that is indistinguishable from causality, within the errors of measurement, in many situations (though not all).
Gelernter states (pg 15) “No memory means no experience”, which contains aspects of truth and aspects that are false and hides many of the levels of complexity present in being human. Conscious memory and emotional memory often use different pathways, and one can be present and the other absent (leading to many pathologies, like Capgras Syndrome or prosopagnosia). Our pattern recognition systems can learn pattern, without having memory of the experiences that trained them, and so carry shadows of that experience (as explored in the Bourne identity novels).
So it isn’t anywhere near that simple. It is vastly more dimensional.
His Fig A, quick sketch, showing a spectrum of focus captures some aspects of awareness, and misses others. In my younger days I trained myself, over many years, to hold my breath a long time (7.5 minutes), and dive deep. Doing that involved many altered states of consciousness and awareness, often with very little awareness. As a student of neurochemistry at the time, and someone prone to deep introspection at the best of times, I find is graphic sort of hints at something, but obfuscates at the same time.
On page 22 he says – “Only the logical argument has predictive power. Only the story has normative moral content.”
Which to me is nonsense.
Logic is a modeling tool. Logic has predictive power within the realm of the model.
All models depend on the constraints that define the boundaries and action sets of both the model and the actors within it.
Where most people seem to be confused, is in the relationship between our experiential reality, and reality itself (whatever reality might actually be).
It seems clear to me that all any of us can experience is our own personal models of reality assembled by our subconscious brains.
This “field of thought” (to borrow a Buddhist expression) is not limited to any experience of reality, and it is usually closely entrained to reality by the inputs from our senses, by the training that past experience has given to our predictor systems, and by any actual memories we have lain down (which will be subject to all sorts of biases resulting from the models and predictors we had at that time, as modulated through the emotional contexts of both past and present, etc).
So it is complex, and his simplistic approach hints at something, but seems to create far more obfuscation than clarity in the quotes above.
If anyone else here has the time and interest to stick with it, I’d be interested in a brief summary of key insights, and an opinion as to if it is worth the effort.