Ezra's Round Table / Systems Seminar: Barry Smith (University at Buffalo)


Frank H. T. Rhodes Hall 253


The Ontology of Complex Systems and the Limits of Artificial Intelligence

The book Why Machines Will Not Rule the World (published August 2022), sets out a theory of systems in general and of complex systems in particular, which we apply in the construction of a detailed account of that feature of human organisms we call intelligence. We can infer from this account an understanding of how an artificial intelligence would need to be structured in order to emulate human intelligence. The upshot is that what AI researchers call 'AGI' (or general artificial intelligence) is impossible, and that AI systems will be forever confined to certain narrow lanes, of which the most conspicuous current examples are the DeepMind Go and the AlphaFold protein structure algorithms.

As an undergraduate, Barry Smith studied mathematics and philosophy at the University of Oxford, before earning his Ph.D. from the University of Manchester in 1976. Currently, he holds the position of Julian Park Distinguished Professor of Philosophy and Adjunct Professor of Biomedical Informatics, Computer Science, and Neurology at the University at Buffalo.

As evidenced from his professorial titles, Smith occupies both the role of philosopher and scientist, blending the two areas of study through his dual focus on ontology and biomedical informatics. Smith has published articles in as many scientific publications as he has in philosophical publications, and his approach can be roughly described as applied ontology.

Smith’s influence is notable outside of academia, such as his involvement with global health organizations in advancing biomedical informatics, and even with the U.S. Army and Air Force.