Panel 1 “Healthcare”
From algorithm hype to clinical practice: using GenAI responsibly
Generative AI (GenAI) is said to have the potential to transform healthcare, public health, and scientific research. Some are even talking about “the AI revolution in healthcare”. While it is imaginable that GenAI could help manage complex medical cases, streamline clinical documentation, or provide timely insights from research, its value has yet to be proven. On the other hand, several immediate risks are already known, such as the generation of inaccurate, incomplete, or even false information; the use of unknown or low-quality training data; or the opacity of red teaming activities. There are also concerns that an overreliance on AI could undermine the skills of healthcare providers. Further at the societal level, there are concerns around the growing dominance of large technology companies, the rising environmental costs of AI development, and the troubling possibility that GenAI could undermine human expertise in critical areas of decision-making.
One approach to address these challenges is participatory design. By actively engaging diverse stakeholders – including patients, clinicians, technologists, and policymakers – participatory design ensures that digital health technologies reflect the needs, concerns, and values of the people they are intended to serve. This approach can help promote AI-driven tools that are equitable and inclusive. But what is needed to bridge the gap between technology development and clinical practice?
This panel brings together experts in healthcare, AI development, ethics, and policy to delve into the complex, dual nature of GenAI in digital health – its ability to bridge gaps in healthcare access and equity, but also its potential to create new divides. The discussion will address ethical issues such as privacy, informed consent, and algorithmic transparency; questions about how AI is changing the role of medical professionals; and broader societal concerns such as disparities in digital literacy and access to AI-driven tools and explore how participatory, interdisciplinary approaches can help address the challenges and opportunities that GenAI presents for digital health.
Panel 2 “Knowledge, Skills and AI”
Knowledge in a digital and AI-enabled world: Which human knowledge and capabilities can AI complement, and help build?
What should humans know and be able to do, what do they have or want to know and be able to to, in order to have a job, to be happy or satisfied? Current technical progress – in generative AI as well as in other areas of computing technology such as robotics – again challenges us to discuss such questions.
The question is naturally asked to provoke: Often, it isn’t even clear yet what modern AI-based systems can do better than humans, be that on their own or with capable humans in the loop. In parallel, humans always have the choice to do and learn whatever they want, to some degree. Nonetheless, the question seems to be: What remains for us humans when AI can do everything – which knowledge and competence is it worth having? In this panel, we address such a concern with a design-oriented question as the starting point:
Which knowledge and capabilities can modern AI complement, and help build?
As the basis and inspiration for HCI research, this panel tries to assess concretely the as-is situation, how modern AI is embedded in different spheres of of life: How does modern AI enhance knowledge (development) and self-efficacy of different people?
This panel thereby connects to overarching societal issues, as well as wider thematic arcs in ongoing Human-Computer Interaction: When are automation, assistive technology, or technology as a tool appropriate, and how can the respective human-computer division of labor be differently designed for? What are key design goals – productivity, self-efficacy and competence, satisfaction or well-being? How can these goals be considered for people with very diverse capabilities, backgrounds and values?