Our weekly SRI Seminar Series welcomes Richard Watson, an associate professor in the Agents, Interaction and Complexity group at the University of Southampton's School of Electronics and Computer Science. Watson has over 80 publications on topics spanning artificial life, robotics, evolutionary computation, population genetics, neural networks, evolutionary theory and computational biology, and is the author of Compositional Evolution: The Impact of Sex, Symbiosis, and Modularity on the Gradualist Framework of Evolution (MIT Press, 2006).
Watson’s research seeks to deepen our understanding of biological evolution by expanding the formal equivalence of learning and evolution—in particular, using connectionist models of cognition and learning. In this talk, he will introduce the concept of “natural induction” as a critique of the ideas of Universal Darwinism that focuses on the evolutionary potential of how organisms develop relationships by working together.
Talk title:
“An antidote to universal Darwinism”
Abstract:
Universal Darwinism is the idea that, regardless of whether we are talking about organisms, corporations, nations or humankind, “survival of the fittest” is the only way that anything can improve. The strong survive and the weak perish—like it or not, that’s nature’s way and the way of the world—so, you may as well take what you can before someone else does and prepare to defend yourself. This mindset is pervasive in Western sociology, contributing to competition, selfishness, isolation, individual and societal extortion, ecological and planetary exploitation—in short, the disintegration of healthy living systems and social well-being. Observationally, we know that healthy natural systems do not work this way: cooperation, mutual benefit and connection is the route to a flourishing ecology and to flourishing social systems. And although adaptive biological complexity might be driven by competition at some level of description, since all complex individuals are composed of parts that used to be individuals themselves, this complexity depends on extraordinary multi-level cooperation and synergistic relationships. Whilst the theory of the competition involved is well developed, the theory of the working-together involved is not. Indeed, the exclusivity assumption of Universal Darwinism—that natural selection is the only possible mechanism of adaptation—forces the conclusion that all cooperation is merely apparent, driven by the selfish motives of the individual or its genes. This exclusivity assumption prevents theoretical development that might better explain the biological complexity we observe, and reinforces exploitative social norms. But is there any alternative? What other natural mechanism of adaptation could there possibly be?
“Natural induction” occurs spontaneously in dynamical systems described by networks of viscoelastic connections experiencing episodic stress. This behaviour follows not from the differential survival or reproduction of components, but from the spontaneous differential easing of frustrated relationships between components (e.g. viscoelastic connections give way under stress). It does not involve the action of natural selection: neither the network as a whole nor the components of the network need to be reproducing entities. It is natural in so much as it does not require a system to be selected or designed for the purpose of exhibiting this behaviour. The effectiveness of this process as a mechanism of adaptation can be quantified in terms of its ability to discover high-quality solutions to difficult optimisation problems (which is different from, and superior to, that of natural selection). The explanation of this ability can be understood through its functional equivalence to connectionist models of cognition, learning, and generalisation. The result can thus be understood as a system that exhibits ‘systemic intelligence,’ holding knowledge from past experience distributed in the organisation of its connections (like a neural network does, and with the same generalisation capabilities), rather than in the frequencies of its particulate constituents (like a naïve genome model does, which cannot exhibit generalisation). Accordingly, natural selection is not the only possible process of adaptation in the natural world. Moreover, the mechanism of natural induction has a lot more to do with finding ways to work-together through the organisation of functional relationships that ease frustrations than with competition and differential survival. This framework offers new explanations for adaptation in biological evolution—in particular, the evolution of evolvability and the evolutionary transitions in individuality—that are difficult to account for under conventional Darwinism.
In this talk, I will introduce natural induction and its implications for Universal Darwinism. I will discuss its implications for how we choose to live as individuals, and, in particular, for how we nurture our relationships with one another and the social systems in which we participate, in order to support systemic intelligence rather than degrade it—so that we may be better equipped to meet the global ecological and social challenges we face.
Suggested readings:
R.A. Watson, E. Szathmáry, “How can evolution learn?” Trends in Ecology & Evolution (February 2016) 31(2): 147–157.
R.A. Watson, C.L. Buckley, R. Mills, “Optimization in “self‐modeling” complex adaptive systems.” Complexity (May/June 2011) 16(5): 17–26.
R.A. Watson, R. Mills, C.L. Buckley, “Global adaptation in networks of selfish components: Emergent associative memory at the system scale.” Artificial Life (2011) 17(3): 147–166.
About Richard Watson
Richard Watson is an associate professor in the Agents, Interaction and Complexity group at the University of Southampton's School of Electronics and Computer Science. He received his BA in Artificial Intelligence from the University of Sussex in 1990 and then worked in industry. Returning to academia, he chose Sussex again for an MSc in knowledge-based systems, where he was introduced to evolutionary modelling. His PhD in computer science at Brandeis University (2002) resulted in 22 publications and a dissertation addressing the algorithmic concepts underlying the major transitions in evolution. A postdoctoral position at Harvard University's Department of Organismic and Evolutionary Biology provided training to complement his computer science background. He now has over 80 journal and conference publications on topics spanning artificial life, robotics, evolutionary computation, population genetics, neural networks, evolutionary theory and computational biology. He is the author of Compositional Evolution: The Impact of Sex, Symbiosis, and Modularity on the Gradualist Framework of Evolution (MIT Press, 2006). His current research seeks to deepen our understanding of biological evolution by expanding the formal equivalence of learning and evolution—in particular, using connectionist models of cognition and learning. A key paper, "How Can Evolution Learn?", with Eors Szathmary, highlighting the core ideas, was published in TREE 2016.
About the SRI Seminar Series
The SRI Seminar Series brings together the Schwartz Reisman community and beyond for a robust exchange of ideas that advance scholarship at the intersection of technology and society. Seminars are led by a leading or emerging scholar and feature extensive discussion.
Each week, a featured speaker will present for 45 minutes, followed by 45 minutes of discussion. Registered attendees will be emailed a Zoom link approximately one hour before the event begins. The event will be recorded and posted online.
Richard Watson