DKR1
At one time the analogy was circulating in computer science that the brain is to the mind as the computer hardware is to the software. This analogy still has implicit attraction for many, tho these days the idea is couched in more respectable academic language.
But, we can safely say that the parallel is false. First of all, there is no fundamental difference between hardware and software. What we have is a hardware system designed to be "universal,' or a general system that can be modified into following specific routines. A software program selects and orders a subset of pre-existent logic gates.
A universal Turing "machine" converts to a specific TM calculation
under the concept that, conceptually, it is possible to choose one computation from the set of all computations. Or, the universal machine uses the standard description (which reduces to an integer) as an input value, and then yields a computation that a separate TM with the same number would yield.
How would one derive consciousness or qualia from Turing's 1936 paper? Yet any known AI system is directly (if only conceptually) convertible into a Turing program. There is no magic added -- that physics knows of -- whereby electronic circuitry implies consciousness. Atheist Turing tried to address this issue with his 1938-9 paper on oracles. He, of course, realized there are one or two things human minds do that computation could never achieve. He proposed a physical solution that limited the scale of the non-standard "objects" that would, he held, correct the undecidability problem. But, it turns out that his physical oracles are not even conceptually possible on account of the Heisenberg uncertainty principle and quantum limits more generally (in that time period, he would not have been expected to know anything much about quantum theory).
Anyone who doubts my case should perhaps re-examine his position, based on the fact that any consistent logic circuit (recognizing that programers sometimes devise inconsistent ones that result in "bugs") can be represented thru use of a single logic gate, such as NAND ("not and").
Consider [A --> B] <--> [~A v B.] <--> [~(A.~B)]
where the dot represents "and."
By this, we see that a logic circuit may be represented as a logical product, with the input values equivalent to an axiom set.
n_Π_(i = 1) = ~(x.~y)_i
The output value is then reached at step n.
A basic hierarchical structure is implicit in the essential recursion present. Other hierarchical structures can be found by, say,
m_Π_(j = a) = ~(x.~y)_j where m<n
Further, of course, we can have more than two consecutive subsets of products, obtaining any level of "complexity."
So for any step 1 to step 2, no qualia are implied. And that holds for any step m to step m+1. But to reach a hierarchical subset, one first has to proceed one step at a time. Hence, no clearcut implication of qualia.
As most of us agree that the consciousness or qualia that constitute it, are human realities, we are again forced to the conclusion that any physics of consciousness is well beyond current conceptions of physical law. After all at present the scientific method assumes deterministic logic, or largely deterministic logic. (The issue of tautology v. contingency is a side issue which is not relevant here.)
+++++++++
TCJ1
It seems quite reasonable to view a theory of physics as a coherent means of description that makes, or attempts to make, accurate predictions. A number of assumptions (whether construed as empirical or not) are, in terms of the theory, axioms or postulates. (Witness Newton and Einstein, for example.)
We see immediately that we have a logical structure underneath the theory, as in:
"If A and B occur close together in time, then C must follow," conceding that each letter can itself represent more than one letter. So we have ~(A.~B) --> C
or ~[(~A.B).C]
We are then free to use the product notation above whereby, interestingly, we have proved that a physical theory is equivalent to a logic circuit, or Turing machine. In fact, all logic circuits, or Turing machines, are the same except for the input value(s). Any consistent physical theory is the same as any other, except for the input values. (This is why Goedel's inconsistency proof focused only on the (conceptual) axioms of number theory.)
We recognize that for a rigorous proof of a scientific claim, we simply use the Deduction Theorem (reverse the progression of the logical product), while accepting that the deduction theorem cannot be used for a set of undecidable physical problems, as we know full well that no modern physical theory can do without number theory (answer to Hilbert question No. 6).
Again we repeat: Any modern theory of physics can be modeled as a logical product, which in turn can be used to represent any logic circuit (or Turing machine). Hence, any purely physical theory of brain functions must be representable as a logic circuit. The many sensory detectors are to be represented as an axiom set, whereby one decision can follow a specific axiom subset.
A train of axiomatic signals "close" in time (a low number of time steps apart). In turn this next-stage decision node may have a time lag, such that other axiom signals hit another next-stage node. These nodes can then be sufficient for a decision. Sometimes a node receives a stream of signals. The stream can be sampled upstream from the decision node and periodically fed back to the decision node as one of the criteria used in the decision-making function.
https://photos.google.com/
photo of exponential graph
The above picture gives a snapshot at an instant of time, rather than the representing the full process. Nevertheless, it shows the simplicity underlying ALL strictly computational complexity.
This is all standard computer science. But the point is that complexity is, in principle, not all that complex. The rubric of "complex" stems from the fact that such systems tend to encompass major areas of unpredictability. But, it does not follow that because complex systems have large swaths of unpredictability, the emergence of the wild card of consciousness is a statistically reasonable idea. Probability is of no use when other considerations militate against such a conclusion.
photo for elsewhere in K notes
at Dece
No comments:
Post a Comment