Greif, Hajo: Ideals of Transparency in Artificial Intelligence and Philosophy. In: Ethical Aspects of AI, P. Stacewicz, Ed., Oficyna Wydawnicza Politechniki Warszawskiej, Warszawa, 2025, pp. 67–73.
This short paper, based on an invited talk at the Ethics and AI conference in Warsaw in September 2024, is the first published manifestation of an inquiry into the notion of transparency in various philosophical subdisciplines. It is well known in contemporary philosophy of science – mostly in discussions of an apparent lack of transparency in AI and computer simulations – but made its first appearance in Anglo-Saxon philosophy in Whitehead/Russel’s Principia Mathematica and Moore’s “Refutation of Idealism”, passing through more recent thinkers like Quine, Dummett, Evans, Dretske and Boghossian. (I don’t discuss most of these authors in this paper, but in the longer treatise I am developing from that material.) Transparency also makes an appearance in theories of the evolution of cognition in the philosophy of biology, explicitly in Sterelny, but also Godfrey-Smith. I task myself with showing that these concepts, although independent and semantically distinct, may jointly help to answer some important questions concerning transparency vs opacity in AI.
The conference volume is Open Access and can be retrieved here (direct pdf download).