Large open mixtures of almost stable states with many transitions, and possibilities
I simply keep track of all states, and then collect the many models, datasets, and estimates of measurements or properties or guesses. When I say “lossless”, I generally mean “simply absorb it all”, verify, try to link to all related things, and “let the computer do most of the work, but never let it run unattended”. If a method has limitations, it generally has a better alternative, though in the world, the crop of alternatives might be so bad, none of them are worthwhile. Ultimately now, the transition if from “writing” that requires human eyes and memories to be in the critical path. And “memory” more generally, that stored the knowledge and “computer” that can process it. As a practical matter, “knowledge” including things like “thermodynamics” can be approached as “everything written”, “all sensors and data”, “all devices and methods”, “all purposes”, “all humans”, and an ideal “reality”.
My own views, albeit based on a lifetime with mostly perfect memory, cannot be audited or verified, until brain scanners and recorders are near perfect. But computers with lots of memory, remembering – classifying – verifying – connecting – indexing – can do much to take over the huge burden in every field of just keeping track of the pieces, parameters, models, parts, names, materials, shapes etc etc etc.
When I was at UT Austin in the 1970’s I hung out with the group of people working with Ilya Prigogine. I do not know if you ever read “Modern Thermodynamics, from Heat Engines to Dissipative Structures” by Ilya Prigogine and Dilip Kondepudi”, It was published in 1998. Now Prigogine’s 1977 Nobel prize in Chemistry was for “dissipative structures” “far from equilibrium”. And with 21 years in between, perhaps it has distilled a few things. Prigogine died in 2003.
The range of things that are “computable” is much larger than those things that can be done with mathematics. And the range of things that can be done with machine intelligence or neural nets, or statistical methods with sensors and control systems is large still. When you have systems where both machines and humans work in close collaboration, most anything is possible now. If you can keep money and greed out.
I appreciate your notes about thermodynamics and scientific methods and ways of thinking. It is nice now and then to think that life can have purpose and meaning, not just existence.
Besides “dissipative systems far from equilibrium”, there are “collaborative and growing systems that can have many stable and productive states”.
Filed as (Large open mixtures of almost stable states with many transitions, and possibilities)
Richard Collins, The Internet Foundation
