Following extensive numerical experiments, Adami [Phys. Lett. A 203 (1
995) 29; Artificial Life 1 (1995) 429] has suggested that the evolutio
n of competing computer programs in artificial life simulations shows
signs of being a self-organized critical process, The primary evidence
for this claim comes from the distribution of the lifetimes of specie
s in the simulations, which appears to follow a power law. We argue th
at, for a number of reasons, it is unlikely that the system is in fact
at a critical point and suggest an alternative explanation for the po
wer-law lifetime distribution. (C) 1997 Published by Elsevier Science
B.V.