Edinburgh Napier University

  A fundamental issue of evolution of life is the emergence and maintenance of self-referential autocatalytic systems (e.g. living cells). In this paper the problem is analyzed from a computational perspective. It is proposed that such systems have to be infinite autocatalytic systems, which can be considered equivalent to Turing machines. The implication of this is that searching for finite autocatalytic systems is likely to not be successful, and any such finite system would be maintainable only in a highly stable environment. The infiniteness of autocatalytic systems also implies that top-down search for the simplest living system is likely to stop at relatively complex cells that are still able to provide a realization of infinite autocatalytic systems.