Exploring the Nuances of Designing (with/for) Artificial Intelligence

TL;DR

This paper discusses the importance of considering infrastructure in AI design to address both technical and societal challenges, emphasizing a holistic approach beyond binary decision-making.

Contribution

It introduces the concept of infrastructure as a framework to integrate algorithmic and societal considerations in AI development.

Findings

Highlights limitations of purely algorithmic solutions

Proposes infrastructure as a holistic design approach

Emphasizes ethical evaluation in AI systems

Abstract

Solutions relying on artificial intelligence are devised to predict data patterns and answer questions that are clearly defined, involve an enumerable set of solutions, clear rules, and inherently binary decision mechanisms. Yet, as they become exponentially implemented in our daily activities, they begin to transcend these initial boundaries and to affect the larger sociotechnical system in which they are situated. In this arrangement, a solution is under pressure to surpass true or false criteria and move to an ethical evaluation of right and wrong. Neither algorithmic solutions, nor purely humanistic ones will be enough to fully mitigate undesirable outcomes in the narrow state of AI or its future incarnations. We must take a holistic view. In this paper we explore the construct of infrastructure as a means to simultaneously address algorithmic and societal issues when designing AI.