Deb referenced this article on my facebook page The Case Against Reality by Donald Hoffman
Some very powerful ideas in that piece, and some that seem to me to be serious errors.
The idea that evolution optimises towards a fitness function is very powerful, and it needs the added notion that it does so by what is essentially a random walk through the possibility spaces available. The optima localisations achieved are local rather than global, which can give some very odd outcomes.
And I am generally aligned with what he says about evolution.
When it comes to his attempts to generalise about consciousness he seems to me to make a couple of very serious (and very common) errors.
One error is around the use of mathematics. Mathematics is a great modeling tool. I use math every day, and have done so for well over 50 years. It is one of the best modeling tools we have, and as such has great utility, and one should never confuse it for the thing it models.
And that could take us down a very deep exploration of epistemology and ontology. And the essence of that path is this sequence.
1 It seems that evolution has tuned us as survival machines with a large set of modeling and heuristic tools in terms of our brains and sensory apparatus and our entire embodied being (with all its many levels of subtle and not so subtle electrochemical interactions). The complexity of a human being is something I have enumerated many times. We are composed of vast numbers of cells (about 10,000 times as many as there are people on this planet), and every cell has about 5 times as many molecules in it as there are cells on our bodies. No chance of ever understanding the detail, we can only ever deal with hugely simplified models.
[Drat – hit enter too early – going to continue working offline, and come back and finish.]
So yes – evolution seems to have gone down the paths it has in our case.
So in that sense, yes, our perceptions, desires, tendencies to distinctions, default model settings, are all tuned in a sense.
And it seems clear that in our case, evolution has gone down a path where generalisation has been favoured, so that we can “tune” all aspects of our being – though only one at a time in a conscious sense, and in another sense, all things effect all other things in some sets of ways – so there is also a connected, almost holographic aspect, to everything we do.
2 It seems that in our case, evolution has delivered systems that produce, by subconscious processes, a model of reality that is slightly predictive in nature – around some 200 milliseconds in most situations, and it can vary quite a bit, depending on context and training.
So, as Hoffman identifies, we experience the model, not reality itself.
3 It seems that mathematics is a great modeling tool, as described above. And about 20 years ago I started to seriously wonder about how useful a tool it really was. So I started looking for some of the simplest mathematical constructs in reality. They don’t seem to exist. I have been unable to find any scale independent examples of any of the simple geometric constructs (like a circle).
So it seems that mathematics can deliver models that are fit for purpose at particular scales, and fail if pushed too far. This understanding was reinforced by Feynman’s sum over life histories approach to QM.
4 In terms of understanding QM and fundamental realities, the collapsing wave function idea is not a required interpretation. There are many alternative ideas that work far better. Rachel Garden’s Global Logic is one approach that retains an essentially Platonic approach to truth but moves to a trinary rather than a binary logic. It works.
And it seems to me that Ockham requires that we go beyond Plato, beyond hard causality, and into deeply stochastic realms.
It seems most likely to me that this reality in which we find ourselves is a very complex system, that at its base has stochastic elements within certain constraints. And once one starts recursively applying complexity theory in this fashion, then all certainty is lost. One is left only with confidence. And in many situations the difference between such confidence and certainty is less than the measurement error on our best instruments.
5 The combination of 50 years interest and study of life, evolution, complexity, and intelligence, combined with 40 years of working with computer systems, both hardware and software, from op code level up through assembly and many levels of languages, working on all levels of the ISO model, has left me with a practical appreciation for complexity and order that is very different from most people.
So I have a certain set of skills at using mathematics, and creating and using modeling tools at many levels, and I attempt to always remain clear that my best models are unlikely to be exactly like the real thing (very close in many aspects perhaps, and probably not at all close in aspects that I have not yet encountered in reality).
So in terms of mathematics, I am a strong yes to using math, it is a very powerful and infinitely deep set of available tools, and it isn’t reality.
The second major error is related.
He takes the step from asserting that because observers are required to collapse wave functions, that consciousness must be an ontological primitive. Which kind of has an internally consistent logic to it, but is not a necessary nor even probable outcome of the full data-sets available.
He used the model of a computer, which is powerful.
I use them a lot as models to think about things.
When one is building complex models, one can create classes with infinitely extensible sets of attributes, and those attributes can have evolving dynamic effects on other agents in the system.
It is quite easy to create systems of sufficient computational complexity that the shortest way to see what they will is to try them and see. Every other way will take longer, if indeed there are any other ways (and for some classes of systems, there are no other ways).
One does not need to assert consciousness as an ontological primitive.
The hard problem of consciousness is basically a failure to be be sufficiently familiar with systems.
At the level of general principles, that problem is resolved for me, and it is a problem of sufficient complexity that it seems the details will be forever beyond any mind – as it seems very probable that the mind is itself the result of those processes.