Information Technology Reference
In-Depth Information
Among the latter there are true reproducers, and after a number of steps these are
the principal structures, as can be seen in Figure 4. Some reproducers suffer variation
in shape or viability due to their interaction with other structures, which indicates that
the robustness induced by the set of rules has a limit.
Fig. 4.
Detail of the result of a successful experiment of the self-inspecting model. Different
copies of true replicators are the main remaining structures after 500 steps.
When self-inspection is present, the rate of success in emerging self-replicating
units is fairly high. Out of the 100 experiments of the self-inspecting model, only in 6
there was no emergence, while in 100% of the combined experiments some kind of
reproducing pattern did appear. This rates are obtained for the mentioned 70% for the
density of quiescent states. Making an emptier or a more crowded environment
reduces significantly these rates, though quantitative measures have not been taken.
A very interesting phenomenon is produced in three experiments. Sometimes a
self-inspecting loop looses its reading signal due to its interaction with surrounding
structures, and so a potential genetic loop is produced, with only construction signals
inside. In these cases the self-inspecting path closes losing its self-inspection signal,
but it doesn't manage to retain its arm an so it cannot further reproduce.
5
Conclusions
We have tried to contribute with experimental evidence to our earlier hypothesis
about the self-inspection strategy as a better adapted one. We have defined a neutral
model in which no bias could be attributable in favor of any of the paradigms.
Additionally, we have advocated for a uniform design, in which the conditions for
emergence would be substantially the same as those in which the emerged patterns
would operate, with the only exception of the changes directly produced in the
cellular space by the new reproducing units.
Our efforts have been directed also toward the design of a model in which the
reproducing entities would be embedded in an unfriendly environment, as opposed to
most reproducing cellular automaton models, which are intended to operate in
laboratory conditions.
