https://techxplore.com/news/2019-10-reservoir-temporal-classification.html This is an interesting development that compliments the previous announcement that a single optical quantum transistor can crush almost any Fast Fourier Transform almost instantly. With this approach they can quickly characterize the Chaotic nature of any system, or the greater context, and how to best apply Fast Fourier Transforms to further refine the content of any models. In other words, combing analog contextual computing with digital computing, which should be possible to do in four rudimentary overlapping ways. In this model, you could add stochastic neuromorphic processing as well, and which one is used for any given calculation would simply depend on what kind of efficiency you require to obtain a decent model. The reservoir they are talking about might as well be described as measuring what's missing from this picture, using higher dimensional mathematics that are humanly inconceivable. Entangled particles can be said to produce the same results with the greatest efficiency possible, and the human brain itself uses both quantum mechanics and classical mechanics, because which is more efficient just depends on the situation. The brain crunches numbers in a fashion similar to a phase transition, walking a drunkards random walk between order and chaos, allowing it to be the most efficient. They've basically invented a crockpot for crunching any data temporally, but that doesn't mean for simple calculations that Newton isn't faster. What it means is the stock market trades will soon be done by computers predicting everything much more accurately than people can. They are talking about using IBM's latest neuromorphic memristors, which they developed at tremendous expense, precisely because they are dirt cheap to make. Until now, nobody has been able to figure out how to program the damned things... Used in something like a video game, it means the computer could easily predict your next move and always stay ahead on frames with 93% accuracy or better when combined with other approaches. A quantum mechanical version could even incorporate entangling the operator and machine for entirely unpredictable emergent effects or results. Their chip is basically a classical version of a quantum simulator or computer. In quantum mechanics, whether a computer is a simulation or just crunching numbers is debatable. Its analog logic and the no-cloning theorem says its just as useful to think of quantum systems as providing simulations of others and measuring them at the same time, because perfect accuracy is impossible. Imitation is the most sincere form of flattery, but its impossible to provide a perfect imitation or clone of information itself, which expresses particle-wave duality.
After reading over the article it occurred to me that this type of computing would ideally involve eight distinctive networks of increasing complexity, that reflect those of the human brain used for pattern matching. You could literally use eight of them to imitate how the human mind works to make decisions, based on pattern matching. So far, half the networks have been clearly identified on one side of the brain, and basically cover the entire side with four circular rings well spaced out. This technology will make it easy to apply the newest AI technology towards imitating a human being. The chips themselves provide the context awareness, in the form of a temporal analysis, while more conventional machine learning can then analyze the context for meaningful content. What they require to perfect the design would be a clear comprehension of the immune system and the four types of consciousness. Autists and schizophrenics can be considered merely a different type of conscious awareness that is related to our immune system and how we interact with the world around us on a cellular level and quantum mechanically. Efficiency is a central issue, and extreme efficiency can be creative, such as one autist who taught himself how to play piano like a concert pianist, in a matter of a few weeks.
