“Informational accounting will reveal that sources of active information are responsible for putatively successful computer-based evolutionary simulations.”
This is closely related to the first prediction, extending it to the realm of computer simulations of evolution as well as the real world of physical processes. The idea is that even a computer program designed to imitate evolution by introducing mutations and a selection algorithm is unable to generate significant complex information without an intelligent agent inserting it in some way.
This prediction is a generalization of the critique that Meyer offers, on pages 283-290, of Ev and Avida, two computer simulation programs that have been published. These programs attempt to model the processes of mutation and natural selection. Meyer’s critique is based on the analysis done by Robert Marks. The claim is that Ev achieves its success largely due to the introduction of so-called “active information” by which Marks means information added by the programmer to reduce the difficulty of the search. The analysis of Avida is more complex. Meyer’s detailed note on pages 533-535 chides Avida more for failing to realistically simulate the generation of new functional genetic information than for supplying too much active information.
It is inevitable that simulating hundreds of millions of years of evolution requires some degree of incorporation of “active information.” Accounting rules for such input are not well established. Settling this prediction is more likely to shed light on the intricacies of simulating real-world situations rather then settling any issues of ID. I invite readers who may have worked with or studied Ev and Avida or similar simulations to comment and enlighten us on more details.