Artificial intelligent (AI) systems that perform image classification tasks are being used to great success in many application contexts. However, many of these systems are opaque, even to experts. This lack of understanding can be problematic for ethical, legal, or practical reasons. The research field Explainable AI (XAI) has therefore developed several approaches to explain image classifiers. The hope is to bring about understanding, e.g., regarding why certain images are classified as belonging to a particular target class. Most of these (...) approaches use visual explanations. Drawing on Elgin’s work (True enough. MIT Press, Cambridge, 2017), I argue that analyzing what those explanations exemplify can help to assess their suitability for producing understanding. More specifically, I suggest to distinguish between two forms of examples according to their suitability for producing understanding. I call these forms samples and exemplars, respectively. Samples are prone to misinterpretation and thus carry the risk of leading to misunderstanding. Exemplars, by contrast, are intentionally designed or chosen to meet contextual requirements and to mitigate the risk of misinterpretation. They are thus preferable for bringing about understanding. By reviewing several XAI approaches directed at image classifiers, I show that most of them explain with samples. If my analysis is correct, it will be beneficial if such explainability methods use explanations that qualify as exemplars. (shrink)

When is it justified to use opaque artificial intelligence (AI) output in medical decision-making? Consideration of this question is of central importance for the responsible use of opaque machine learning (ML) models, which have been shown to produce accurate and reliable diagnoses, prognoses, and treatment suggestions in medicine. In this article, I discuss the merits of two answers to the question. According to the Explanation View, clinicians must have access to an explanation of why an output was produced. According to (...) the Validation View, it is sufficient that the AI system has been validated using established standards for safety and reliability. I defend the Explanation View against two lines of criticism, and I argue that within the framework of evidence-based medicine mere validation seems insufficient for the use of AI output. I end by characterizing the epistemic responsibility of clinicians and point out how a mere AI output cannot in itself ground a practical conclusion about what to do. (shrink)

The ongoing explosion of interest in artificial intelligence is fueled in part by recently developed techniques in machine learning. Those techniques allow automated systems to process huge amounts of data, utilizing mathematical methods that depart from traditional statistical approaches, and resulting in impressive advancements in our ability to make predictions and uncover correlations across a host of interesting domains. But as is now widely discussed, the way that those systems arrive at their outputs is often opaque, even to the experts (...) who design and deploy them. Is it morally problematic to make use of opaque automated methods when making high-stakes decisions, like whether to issue a loan to an applicant, or whether to approve a parole request? Many scholars answer in the affirmative. However, there is no widely accepted explanation for why transparent systems are morally preferable to opaque systems. We argue that the use of automated decision-making systems sometimes violates duties of consideration that are owed by decision-makers to decision-subjects, duties that are both epistemic and practical in character. Violations of that kind generate a weighty consideration against the use of opaque decision systems. In the course of defending our approach, we show that it is able to address three major challenges sometimes leveled against attempts to defend the moral import of transparency in automated decision-making. (shrink)

In this paper I argue that, given the way that AI models work and the way that ordinary human rationality works, it is very likely that people are anthropomorphizing AI, with potentially serious consequences. I start with the core idea, recently defended by Thomas Kelly (2022) among others, that bias involves a systematic departure from a genuine standard or norm. I briefly discuss how bias can take on different explicit, implicit, and “truly implicit” (Johnson 2021) forms such as bias by (...) proxy. I then discuss biased anthropomorphism of technology, focusing on the case of Large Language Models (LLMs) like chatGPT. As with other kinds of bias, there are importantly different kinds of anthropomorphism, some of which can persist without others, and some of which can encourage others. Anthropomorphism can take rather subtle, implicit forms, that can be difficult to detect, resist, and dislodge. Attention to these kinds of anthropomorphism can help us avoid confusing importantly different kinds of evaluation, better assess the risks, and inform strategies for bias prevention and mitigation. (shrink)

The stakes associated with an algorithmic decision are often said to play a role in determining whether the decision engenders a right to an explanation. More specifically, “high stakes” decisions are often said to engender such a right to explanation whereas “low stakes” or “non-high” stakes decisions do not. While the overall gist of these ideas is clear enough, the details are lacking. In this paper, we aim to provide these details through a detailed investigation of what we will call (...) the “Simple Stakes Thesis”. The Simple Stakes Thesis, as it will turn out, is too simple. For even if the stakes associated with a specific one-off decision are low—and hence does not engender a right to an explanation—such decisions may nevertheless form part of a high stakes pattern or aggregate of decisions. In such cases, we argue, even a low stakes decision may engender a right to explanation. Not only does this show that the right to explanation is more demanding than so far recognized. It also shows that the stakes thesis is significantly harder to apply in practice. (shrink)

An important module in the Belief-Desire-Intention architecture for artificial agents (which builds on Michael Bratman's work in the philosophy of action) focuses on the task of intention reconsideration. The theoretical task is to formulate principles governing when an agent ought to undo a prior committed intention and reopen deliberation. Extant proposals for such a principle, if sufficiently detailed, are either too task-specific or too computationally demanding. I propose that an agent ought to reconsider an intention whenever some incompatible prospect is (...) sufficiently valuable along some dimension that can be assessed at zero or near-zero computational cost. (shrink)

I argue that ML models used in science function as highly idealized toy models. If we treat ML models as a type of highly idealized toy model, then we can deploy standard representational and epistemic strategies from the toy model literature to explain why ML models can still provide epistemic success despite their lack of similarity to their targets.