AI-Management Decision Making: Explanation and Transparency


Image: cocoparisienne (Pixabay)


One of the challenges we face in AI-Management Decision Making is how we can make decisions more transparent. Transparency is needed to build trust in these systems and ensure good governance. A suitable method of explaining decisions using knowledge of causes can be borrowed from ancient Greek philosopher Aristotle. He proposes we need to address four questions where possible for a better explanation: what it is, what it is made of, what brought it about, and what it is for (see below). Aristotle says these four causes may not always be used in explanations, for example, in the case of coincidences, which lack final causes, or mathematical abstractions, which lack materials causes (a problem many AI algorithms face). But good scientific decision-making (which we tend to use also in business or management decisions) should be able to answer all four of the questions given below. I have applied it to AI and am happy to get feedback from you and more examples.