News

Critical Perspectives are Crucial for Future Better AI

Published: November 30, 2022

Participants at the AI for Humanity and Society 2023 workshops. Photo: Charlotte Carlberg

Attending to how power shifts between systems and life goals, the redistribution of tasks between different groups, and attending to datasets that reflect the many biases of society were some of the topics discussed at the WASP-HS conference AI for Humanity and Society 2022. Over 350 researchers, politicians, and representatives from the business sector met at Grand Hôtel in Stockholm on November 17 to discuss the next steps for our common future with AI.

As life and its many aspects is complex, so is welcoming, and integrating, AI systems to it. Even more complex does it get as the AI systems must attend to many different types of lives as it is integrated in our widespread and ever-changing society. At this year’s WASP-HS conference, AI for Humanity and Society, it was made it clear that there are many challenges with integrating AI in society but also that the move for just and responsible AI is strong.

“Throughout this conference I believe that it has been made even more clear that there is a space for utopianism for the AI-world. There is a lot of money, hope, and energy going into producing better AI for all. As we do this, social, humanistic, and critical perspectives of AI are crucial” says Francis Lee, Associate Professor of Technology and Social Change, Chalmers University of Technology.

200 researchers, politicians, and representatives from the business sector met on-site at Grand Hôtel for the WASP-HS conference, AI for Humanity and Society 2022, on November 17. An additional 160 persons attended the conference online. The day started with a welcoming and introduction by Anna-Sara Lind, Professor of Public Law, Uppsala University, Francis Lee, Associate Professor, Technology and Society, Chalmers University of Technology, and  Virginia Dignum, Professor in Responsible AI, Program Director of WASP-HS, Umeå University.

After the introduction Martin Ulbrich, Senior Expert in the AI Policy Unit of the European Commission gave a keynote speech on tomorrow’s AI politics and which roads will follow. Carolina Brånby, Digital Policy Director at The Confederation of Swedish Enterprise (Svenskt Näringsliv) added to the keynote as discussant.

After the keynote followed an interactive discussion session of the mapping the grand social challenges of AI. The WASP-HS researchers who gave presentations were; Johan Wejryd, PostDoc in Political Science, Institute for Housing and Urban research, Marisa Ponti, Associate Professor in Informatics, University of Gothenburg, Teresa Ceratto-Pargman, Professor of Human-Machine Interaction, Stockholm University, Pedro Sanches, Assistant Professor in Human-Centered AI, Umeå University, Ericka Johnson, Professor in Technology and Social Change, Linköping University, and Jason Tucker, Associate Professsor in the Global Politics of AI and Health, Malmö University.

“AI is the dream of understanding and reproducing human intelligence in artificial bodies. This dream is not new, but it keeps reinventing itself, and AI becomes a mirror where we project our hopes and anxieties about the future. Someone’s dream may be another person’s nightmare, and even if hopes and anxieties around AI are not always realistic, depending on your perspective it is legitimate to be afraid of being replaced, or bureaucracies becoming more obscure. The general discourse in AI lacks perspective, both in terms of understanding technological adoption over time, and in terms of who gets to dream about AI futures” says Pedro Sanches, Assistant Professor in Human-Centered AI, Umeå University.

Different perspectives were represented at the conference both among the participants and speakers. Abeba Birhane, Senior Fellow in Trustworthy AI at Mozilla Foundation, did for instance present a keynote speech on “AI for All Humanity?”.

“At this conference, there has been plenty of discussions about long-termism, which unfortunately is a hypothetical First World Problem, removed from concrete realities of current AI. I think it is important to understand and tackle discrimination, biases and injustice that are features of today’s AI systems. In my view we should prioritize our efforts, findings, and time on these real current issues” says Abeba Birhane, Senior Fellow in Trustworthy AI at Mozilla Foundation.

Birhane’s keynote was added upon by Jeroen van den Hoven, Professor of Ethics and Technology, Delft University of Technology, Maria Marouda, Chair, European Commission against Racism and Intolerance, and Jennifer Rhee, Associate Professor at the Department of English at Virginia Commonwealth University as discussants.

The day ended with a panel discussion between Pernilla Johnsson, Ericsson Research,
Anders Sandberg, Senior Research Fellow, Future of Humanity Institute, University of Oxford
Katherine Harrison, Associate Professor, Docent, Department of Thematic Studies (TEMA), Linköping University, Andrew McAfee, Principal Research Scientist, MIT Sloan School of Management, on the next step for social sciences and the humanities in AI.

During the conference, it was illustrated through both keynote speeches and discussions that this is only the beginning of the discussion on all the perspectives needed to integrate AI for the benefit of our common future.