System Analysis for Responsible Design of Modern AI/ML Systems
This addresses the issue of irresponsible AI/ML deployment for practitioners and designers, but it appears incremental as it applies an existing system analysis framework to ML.
The paper tackles the problem of irresponsible use of ML algorithms by proposing that a system analysis perspective is needed for designing and implementing ML systems, and it reviews how this methodology can formally evaluate and enable responsible ML practices.
The irresponsible use of ML algorithms in practical settings has received a lot of deserved attention in the recent years. We posit that the traditional system analysis perspective is needed when designing and implementing ML algorithms and systems. Such perspective can provide a formal way for evaluating and enabling responsible ML practices. In this paper, we review components of the System Analysis methodology and highlight how they connect and enable responsible practices of ML design.