Re: AI approach couples biological computation with value



On 19 Gen, 17:54, "Alpha" <OmegaZero2...@xxxxxxxxx> wrote:
Read Montague's new book: "Why Choose This Book: How We Make Decisions",
contains many gems. One of which, and the focus of the book is:

"Nature has equipped biological computations with a measure of their value."

Read is professor of neuroscience at Baylor College of medicine, Director of
the Neuroimaging Lab, director of the Cente for Theoretical Neuroscience and
a fellow at the Institute for Advanced Study at Princeton.

Read visits the COTM (Computational Theory of Mind - if you do not know what
this refers to, in a nutshell, it is that mind is software running on the
brain; and that the mind is not equivalent to brain just as Microsoft Word
is not the logic gates that it runs on) and adheres to its tenets, along
with Turings ideas on computation. However...

But the machine (the logic gates) Word is running on can be emulated in
software, and Word can be in principle implemented in hardware,
therefore, formally, there is no difference.

Unlike computers,"... biological computations are not lifeless streams of
symbols, totally devoid of meaning [*]. Instead, biological computation
carry something extra - an extra measure of their overall worth. Instead of
just computation, there is computation plus something else, and that
something else is a measure of the value of the computation to the overall
success of the organism, its overall fitness. Biological computtions know
how to care"

This looks like a loose definition of reinforcement learning, which
*is* symbolic processing.

He goes on to tell how brain implements goals, and how the energy efficiency
of a computation, selected via evolutionary mechanisms, puts meaning into
neural computations, assigning values and passing the values from one symbol
to another. Note that energy allocation schemes are missing from computers
today (which is one reason why your brain remains cool and your computer's
CPU is hot to the touch! All computations in today's computers get equal
energy billing.) Not so with biological computations, which are almost
"miraculously efficient", WRT energy Read says.

Actually energy efficiency is a major consideration in mobile devices.

* Note that this observation and his subsequent approach, dovetails with the
philosophical position that a computer based on todays technology (without
some advance along the lines Read suggests) will continue to *NOT*
understand.

I just finished the first chapter and look forwrd to the rest of this
excellent (so far) book.

--
Posted via a free Usenet account fromhttp://www.teranews.com

.