The properties of most matter in the universe can be
approximated using
continuum mechanics.
This
is an astonishing simplification since it allows us to ignore all of the
individual atoms and molecules of materials, fluids and gases and just consider
them as continuous materials with well defined properties.
For instance, a grain of salt contains about 1.2 X 1018
atoms, yet since it has cubic symmetry; its linear elastic properties are
completely defined by just 3 numbers.
Most of the technologies mankind has developed so far
exploit this massive simplification to the hilt. Materials scientists (like me) design new
materials with different properties which engineers then put to use to make new
designs and technologies, all the while ignoring the atomic complexity of those
materials.
However, there is a small subset of materials in the
universe that have properties that can not be approximated but rather are complex
and functional at the molecular level. These
materials are made up of atoms that have been temporarily scooped up by the
Darwinian algorithm and bonded with other atoms in precise ways to make tiny
little machines that have precise functions within the context of living cells
or organisms. All molecular machines have specific functions that they perform which can not be generalized to other
molecular machines. Therefore, there is
not a broad simplification that can be applied to all the machines that will
simplify our task in understanding how they work.
Since science is always reductionistic, the second class of
materials in the universe can not be easily understood by science unlike the
first class of materials described above.
In a sense, Mother Nature temporarily infuses the second class of
materials with a sort of magic power I call Soft Vitalism. Soft because it is only ‘sort of’ magic. The second class of materials is all built on
hard naturalistic ground. I am not arguing
for old-fashioned Vitalism which said that there is a supernatural life force
that animates living material. I am only saying that
Darwinism produces material that is mostly resistant to the inroads of science.
This is why I am skeptical about Transhumanism. So far, most of the technologies that we have
made have used the first type of materials which are amendable to reductionistic
science. However, Transhumanists believe
that science will be just as successful with the second class of materials as
it has been with the first.
I doubt it.
Excellent point. I think ultimately transhumanism hinges on the idea of informational redundancy - that we can effectively ignore vast amounts of information and still achieve a given utility function, or some desired or interesting effect - essentially (though not precisely), continuum mechanics. Kurzweil talks a lot about this in his most recent book (for obvious reasons), and he calls it complexity. To use his example, a common rock is made up of perhaps sextillions of atoms, but for functional purposes, we can describe it in far less informationally dense terms because most of the information in that rock is irrelevant to the specific aspects of that rock that we are interested in. If we wish to describe the atomic structure of a rock, we can describe a few of the atoms or molecules, and then, basically, include an appendage reading, roughly, “… repeat n*sextillion times, including variations within the following parameters..." Or we can really describe the rock at any resolution, in relatively simple terms (for instance, by describing the proportions of different elements contributing to the composition of the rock), as long as we do not try to describe the rock at all resolutions (which really would require many sextillions of bits of information). So, I would argue (as Kurzweil does also) that transhumanism is largely dependant for its effective goals on the idea that we do not need to describe every last detail of some thing to do that thing.
Clearly, we can already do, more or less, what the human arm can do, in terms of lifting and moving (though not in terms of manual dexterity), with robotic arms despite robotic arms having vastly fewer functional components than human arms, because we know generally what the arm is supposed to do, and so we can achieve the high level effect without achieving all of the low level ones. The question then becomes, to what resolution must we describe, say, the brain in order to build an effective, functional facsimile of it? Do we need to know the position of every neurone? Every synapse? Every neurotransmitter? Every molecule, atom, or quark? How much of this can be effectively ignored? What is the greatest functional complexity of the human “mind” (NOT of the human brain), and can we define a limit to that function? So, yes, neurones are staggeringly complex computational substrates, but is ALL of this function necessary to approximate a mind? This is one area among many where the transhumanist hypothesis clearly waits (to quote the 2005 Technology Review SENS Challange jury) "possibly in vain, for independent verification." Much of the nascent field of synthetic biology is relative to the hijacking of living mechanisms to effect an artificial function, and this may provide a severely imperfect "out" to transhumanism for a negative on this verification, but if this is the case, we might have to wait for QUITE a long time to see even the most primordial of transhumanist ambitions to come to fruition.
Reply to this