Property of Innovation Under Different Group Size in an Open-ended Fitness Landscape: Demonstration Through Evolutionary Logical Circuits

02/03/2020 ∙ by Masaki Suyama, et al. ∙ 0

Open-ended fitness landscape or architectural innovation is a key characteristic of combinatorial technological evolution. Though many have argued that this feature is important, many models were created in a closed fitness landscape. In this article, we modified a simulation by Arthur and Polak (2006) that used a computer simulation that builds logical circuits with circuits that were built in earlier trials. We used this simulation to investigate (1) whether the speed of innovation is increased by increasing group size and (2) if the nature of innovation would differ when innovation was separated by invention (new innovation that serves a purpose that never was used before) and improvement (already made innovation but is more efficient). The results indicated that indeed group size increases the speed of innovations but is limited than expected. Also, when innovation was separated with invention and improvement, the nature of the two differed. In improvement, the trajectory followed a convex function with productivity of one agent decreasing as group size increased. In invention, the trajectory showed a continuous pattern of rapid increase followed by a plateau. In sum, the results indicated that group size does increase both innovation and improvement, but the effect diminishes as group size increase in an open-ended fitness landscape and when innovation was split, invention resembled a technology s-curve.



There are no comments yet.


page 1

page 2

page 3

page 4

This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.