The AI field needs translational Ethical AI research

Abstract

Calls for Ethical AI have become urgent and pervasive, especially as ethical issues surrounding AI products at tech companies are increasingly scrutinized by the public. Yet even after a first wave of responses to these calls coalesced around Ethical AI principles to guide decision-making and a second wave generated technical tools to mitigate specific ethical issues, multiple lines of evidence indicate that these Ethical AI principles and technical tools have only a limited impact on the daily practices of AI users and producers. In other words, there is a big gap between what we publish in academic papers and what AI creators need to generate AI products that reflect society's values. Ethical AI is by no means the only field to have this problem. However, when medical and ecology fields documented similar gaps between their fields' scientific discoveries and the practices and products that people actually use, they invested tremendous resources into subfields that developed evidence about how to translate what was done in the lab to adopted solutions. I argue in this commentary that it is our research community's moral duty to invest in our own subfield of "Translational Ethical AI" that will determine how best to ensure AI practitioners can implement the Ethical AI technical tools we publish in academic venues in production settings. Further, I offer concrete steps for doing that, drawing on insights gleaned from other translational fields. Closing the "Ethical AI Publication-to-Practice gap" will be a considerable transdisciplinary challenge, but one of the AI research community has the unique expertise, political leverage, and moral responsibility to tackle.

Sync fast and solve things—best practices for responsible digital health

“…Further, even when institutions are aware of responsibility, their lack of resources and funding constitute obstacles to responsible innovation practice 8 . Beyond awareness, it has also been shown that practical translation of ethics into coding and design practices, and business models, is required to ensure that innovation creates real societal value 16 .…”

“…While these can be a vital translational force, commentators have warned against reducing ethical problems in technology (e.g., privacy) to techno-solutionism, the belief that technology alone will adequately solve major societal and political problems 53 . As stated by Borg 16 “technical ethical AI tools are necessary, even if insufficient, for addressing the ethical issues AI poses”. Going forward, we expect the need for ethical reflection and responsibility to remain relevant at all levels of technology evolution.…”

“…Technical solutions to further responsible digital health innovation have also emerged 13 . However, these tools and approaches have not achieved broad adoption and it remains questionable whether they correctly and sufficiently address practical impediments to responsible innovation 14 – 16 . Most approaches, for instance, only focus on parts of the complex stakeholder ecosystem and innovation process, while power and responsibility are disseminated across the ecosystem 2 , 17 .…”

“Sync fast and solve things” – best-practices for responsible digital health

“…Technical solutions to further responsible digital health innovation have also emerged (Morley et al 2020). However, these tools and approaches have not achieved broad adoption and it remains questionable whether they correctly and su ciently address practical impediments to responsible innovation (Schiff et al 2020;Vakkuri et al 2020;Borg 2022). Most approaches, for instance, only focus on parts of the complex stakeholder ecosystem and innovation process, while power and responsibility are widely disseminated (Trocin et al 2021;Martin 2019).…”

“…Beyond awareness, it has also been shown that more "hard" (i.e. practical) translation of ethics into coding and design practices, and business models, is required to ensure that innovation creates real societal value(Borg 2022).Su ciently incentivizing stakeholders to drive such "innovation of innovation" is vital to attaining responsible digital health. Our participants re ected extensively on the role of startups and their dependency on VC funds and larger corporate players.…”