Specific versus unspecific attachment
click to enlarge
This distribution of control and memory has several consequences:
1.
The execution time of a certain program increases because
building up the connections between the many micro-controllers takes time.
2.
Because of the recognition-procedure, absolute addressing
becomes the exception. Long very-specific recognition sites are evolutionarily
costly and sensitive to noise. Even with long recognition sites, copies of the
molecules or micro-controllers will still exhibit exactly the same recognition
site. This is a fundamental difference to standard programming. If a
memory-location is copied, the address of the copy is per definition different
than the source. This means:
a)
There is no longer a clear distinction between copied code and
its template.
b)
When calling a subroutine, the caller has no assurance that
the subroutine exhibiting this recognition-site is indeed the one the caller
wants to execute.
c)
If such a subroutine has been destroyed or corrupted, a chance
still exists that a neighboring intact copy will be executed.
3.
Even if a large fraction of the system is destroyed, the
distributed algorithm might still function and the code survive in an
evolutionary context.
4.
Of course, with this distributed approach, the door is wide
open for parasites e.g. viruses and exploiters of other kinds. This might even
be an advantage for evolution, because parasites not only exploit the system
but they also provide new information to the system and enable it to react more
flexibly.
5.
Recognition as such must be tunable, because, as can be seen in the picture (see above),
without unspecific attachment the replication cycles remain small and sensitive
to perturbations. A proliferation of successful replicates is then not
possible. Switching between specific and unspecific recognition can be realized
with different recognition-sites.
6.
Several recognition-sites per molecule or micro-controller in
addition allow for execution-starts in-middle of the programs (cf TATA-boxes in
the interpretation of genetic information) which helps to simulate
return-entry-points in the micro-controller-programs.