TY - JOUR
T1 - A biological perspective on evolutionary computation
AU - Miikkulainen, Risto
AU - Forrest, Stephanie
N1 - Funding Information:
We thank R. Axelrod, D. Erwin, A. Graham and M. Mitchell for their suggestions on an earlier version of this paper, and R. Lenski and M. Lynch for many helpful discussions on evolution. Thanks to P. Reiter, F. Gomez and C. Schoolland for the original drawings for Fig. 1a–c. R.M. was partially supported by NSF DBI-0939454, DARPA FA8750-18-C-0103 and HR0011-18-2-0024, and NIH 1U01DC014922; S.F. was partially supported by NSF CCF-1908633 and IOS 2029696, DARPA FA8750-19C-0003 and N6600120C4020, and AFRL FA8750-19-1-0501; both R.M. and S.F. were supported by NSF IIS-2020103.
Publisher Copyright:
© 2021, Springer Nature Limited.
PY - 2021/1
Y1 - 2021/1
N2 - Evolutionary computation is inspired by the mechanisms of biological evolution. With algorithmic improvements and increasing computing resources, evolutionary computation has discovered creative and innovative solutions to challenging practical problems. This paper evaluates how today’s evolutionary computation compares to biological evolution and how it may fall short. A small number of well-accepted characteristics of biological evolution are considered: openendedness, major transitions in organizational structure, neutrality and genetic drift, multi-objectivity, complex genotype-to-phenotype mappings and co-evolution. Evolutionary computation exhibits many of these to some extent but more can be achieved by scaling up with available computing and by emulating biology more carefully. In particular, evolutionary computation diverges from biological evolution in three key respects: it is based on small populations and strong selection; it typically uses direct genotype-to-phenotype mappings; and it does not achieve major organizational transitions. These shortcomings suggest a roadmap for future evolutionary computation research, and point to gaps in our understanding of how biology discovers major transitions. Advances in these areas can lead to evolutionary computation that approaches the complexity and flexibility of biology, and can serve as an executable model of biological processes.
AB - Evolutionary computation is inspired by the mechanisms of biological evolution. With algorithmic improvements and increasing computing resources, evolutionary computation has discovered creative and innovative solutions to challenging practical problems. This paper evaluates how today’s evolutionary computation compares to biological evolution and how it may fall short. A small number of well-accepted characteristics of biological evolution are considered: openendedness, major transitions in organizational structure, neutrality and genetic drift, multi-objectivity, complex genotype-to-phenotype mappings and co-evolution. Evolutionary computation exhibits many of these to some extent but more can be achieved by scaling up with available computing and by emulating biology more carefully. In particular, evolutionary computation diverges from biological evolution in three key respects: it is based on small populations and strong selection; it typically uses direct genotype-to-phenotype mappings; and it does not achieve major organizational transitions. These shortcomings suggest a roadmap for future evolutionary computation research, and point to gaps in our understanding of how biology discovers major transitions. Advances in these areas can lead to evolutionary computation that approaches the complexity and flexibility of biology, and can serve as an executable model of biological processes.
UR - http://www.scopus.com/inward/record.url?scp=85100106574&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85100106574&partnerID=8YFLogxK
U2 - 10.1038/s42256-020-00278-8
DO - 10.1038/s42256-020-00278-8
M3 - Article
AN - SCOPUS:85100106574
SN - 2522-5839
VL - 3
SP - 9
EP - 15
JO - Nature Machine Intelligence
JF - Nature Machine Intelligence
IS - 1
ER -