Readers will recognize this theme, and this page from Journal of Open, Distance, and Digital Education leads with my article, On ethical AI principles, and then follows up with criticisms from Jon Dron, Stella George, and another from Dagmar Monett. My target is work by Luciano Floridi (such as) suggesting there is a global consensus on ethical principles, and also the many policy statements, guidelines, and even laws, entrenching them into practice. A closer look reveals no such principles are anything like universal, and some (like explainability and accountability) aren't even ethical principles at all. The Dron and George response is (on my reading) essentially the assertion that, yes, principles aren;t universal; "principles are foundational guidelines, starting points, and orientations that are used to frame understanding and assist with decisions." Monett criticizes my characterization of AI: "when defining AI, introducing a new definition and not considering at least one of the many that already exist... is a questionable omission.... "reviewing, summarizing, translating and composing" are overstatements of the capabilities of AI algorithms." Image: Cogent Infotech.
Today: Total: [Direct link] [Share]
