Why AI is not an Epistemic Authority Regarding Our Own Minds
摘要
Casey Doyle has recently argued that AI is not an epistemic authority for us regarding what beliefs, desires, or intentions we have, so we should not defer to AI about our own minds. Central to Doyle’s argument is his view about what is necessary for a source of information to be authoritative. In this paper, I reject Doyle’s necessary condition for epistemic authorities and propose a new one. According to the condition I propose, AI is still ruled out as an epistemic authority for us when we inquire into our own minds. The newly proposed condition also entails that AI is not an epistemic authority for others when they inquire into our minds, thereby helping to provide a more comprehensive answer to a central question discussed in Doyle’s article.