How do we build AI applications that people can trust?
On 13 September, a group of expert panellists examined this complex question in the inaugural discussion of the Human+ Tech Talks series.
The Human+ research fellowship programme at Trinity College Dublin (TCD) connects Computer Science researchers with researchers from the Arts and Humanities and Enterprise to forge a human-centric approach to technology development. The programme is co-funded by the European Commission’s Marie Sklodowska-Curie actions and led by the Trinity Long Room Hub Arts and Humanities Research Institute and ADAPT, the world-leading SFI Research Centre for AI-Driven Digital Content Technology at TCD.
The Human+ Tech Talks series showcases the research topics of the Human+ Fellows as they discuss prescient technological questions with multi-disciplinary teams of supervisors, enterprise mentors, and other key specialists in this area.
On Tuesday, September 13th, Vincent Wade, co-Principal Investigator of the Human+ Programme and Director of ADAPT, was joined by Linda Hogan, Professor of Ecumenics TCD; Adam Henschke, Associate Professor of Philosophy, University of Twente; and David Lewis, Associate Professor of Computer Science, TCD. The panel discussed the ethical dimension of human-centric AI.
“AI really should be focused on empowering individuals, empowering humans and society to engage in digital experiences,” Vincent said as he kicked off the conversation.
Linda Hogan discussed some of the human values that should be encoded in AI, and the developing tools that can be used to help educate designers. “We already have broad agreement on values and principles of ethical AI. We already have the frameworks, so let’s use them,” Linda said.
Coming from a philosophical and ethicist perspective, Adam Henschke explored issues of agency, and the increase in responsibility of those designing, developing, using, and procuring AI as the field becomes more transparent. “As we gain more understanding of how AI might be used, its applications, and contexts, we’re probably going to see more risks and challenges coming from that,” Adam said.
To conclude the presentations, David Lewis spoke about putting AI principles into practice from a technical standpoint. “We’re in this place where suddenly the complexity of what we need to consider goes beyond an ethical reasoning approach and we have to look at all these other legislations that are out there,” David said.
The next Human+ Tech Talk is on Tuesday, 27 September, as Human+ fellow Kata Szita and her expert panel will explore digital avatars, identity and behaviour. You can register for the event here.
To learn more about these research projects, you can read profiles of the fellows in the Human+ interview series.