This is my latest publication, an adaptation of a post I authored last summer. It makes the argument that we cannot base AI ethics (or any other ethics) on a set of commonly-held principles, for several reasons: first, because the principles themselves are too vague to apply; second, because there is not unanimity of support for these principles; and third, even organizations that support these principles are quite willing to dispense with them in particular circumstances. The paper could be better-written (and was improved a lot thanks to the heroic efforts of the editor) and I really do hope the reviewers come out with their rejoinder, which would also be a valuable addition to the discussion. I should also add that my employer wants nothing to do with this paper and does not want to be associated with it in any way.
Today: Total: [Direct link] [Share]
