Beyond transparency: computational reliabilism as an externalist epistemology of algorithms
Juan Manuel Durán
TL;DR
The chapter addresses how to justify outputs from epistemically opaque algorithms in scientific practice and argues that traditional internalist notions of transparency are insufficient. It introduces computational reliabilism ($CR$), an externalist epistemology that grounds justification in diverse Reliability Indicators ($Type_1$-$RI$, $Type_2$-$RI$, $Type_3$-$RI$) and their tokens, rather than inner algorithmic mechanisms. By emphasizing technical performance, domain-embedded scientific practice, and social coherence, CR reframes justification as grounded in the algorithm’s reliability across specification, coding, use, and maintenance, acknowledging context-dependence and provisionality. Through case discussions around BenevolentAI and Wu & Zhang, the chapter illustrates how reliability indicators—together with expert knowledge and scientific debate—provide a more robust justification for algorithmic outputs than transparency alone, with implications for domain-specific deployment and ongoing methodological refinement.
Abstract
This chapter is interested in the epistemology of algorithms. As I intend to approach the topic, this is an issue about epistemic justification. Current approaches to justification emphasize the transparency of algorithms, which entails elucidating their internal mechanisms -- such as functions and variables -- and demonstrating how (or that) these produce outputs. Thus, the mode of justification through transparency is contingent on what can be shown about the algorithm and, in this sense, is internal to the algorithm. In contrast, I advocate for an externalist epistemology of algorithms that I term computational reliabilism (CR). While I have previously introduced and examined CR in the field of computer simulations ([42, 53, 4]), this chapter extends this reliabilist epistemology to encompass a broader spectrum of algorithms utilized in various scientific disciplines, with a particular emphasis on machine learning applications. At its core, CR posits that an algorithm's output is justified if it is produced by a reliable algorithm. A reliable algorithm is one that has been specified, coded, used, and maintained utilizing reliability indicators. These reliability indicators stem from formal methods, algorithmic metrics, expert competencies, cultures of research, and other scientific endeavors. The primary aim of this chapter is to delineate the foundations of CR, explicate its operational mechanisms, and outline its potential as an externalist epistemology of algorithms.