Zoom: https://umac.zoom.us/j/97781274783 Password: 869095   Abstract There is now a very extensive debate, both inside and outside academia, about how to make very advanced AI systems safe and ethical. Sometimes this is called the Alignment Problem. There are many proposals for how this can be done. In this talk, I argue that most (or all) of these efforts are both futile and potentially dangerous. I end with some old-fashioned suggestions for how to move forward when thinking about AI risk.   Bio Herman Cappelen is Chair Professor of philosophy at the University of Hong Kong. Before moving to Hong Kong, he worked at the Universities of Oslo, St. Andrews, Oxford, and Vassar College. To name just some of his accolades, […]

Abstract The classical Greeks give us a substance ontology grounded in “being qua being” or “being per se” (to on he on) that guarantees a permanent and unchanging subject as the substratum for the human experience. With the combination of eidos and telos as the formal and final cause of independent things such as persons, this “sub-stance” necessarily persists through change. This substratum or essence includes its purpose for being, and is defining of the “what-it-means-to-be-a-thing-of-this-kind” of any particular thing in setting a closed, exclusive boundary and the strict identity necessary for it to be this, and not that. In the Yijing 易經 or Book of Changes we find a vocabulary that makes explicit cosmological assumptions that are a stark […]