For its 772nd luncheon at the Hotel d’Angleterre on September 17, 2019, the American International Club hosted a panel on Artificial Intelligence and Genome Science and its Applications in Medicine and Public Health Fields. Composing the panel were:
Emmanouil (Manolis) Dermitzakis, Professor of Genetics, University of Geneva Medical School, Director, Health 2030 Genome Center
Antoine Geissbuhler, Chief-physician, division of e-Health and telemedicine, vice-rector University of Geneva
Prof. Marcel Salathé, School of Life Sciences, School of Computer and Communication Sciences; Academic Director, Extension School, EPFL
The panel discussion was led by Serge Michel, Editorial Director of Heidi News.
We often hear about genomic science and genetic tests, but these topics remain difficult to grasp for a layman. The AIC luncheon event brought together three experts who have been working in these fields for years. They gave us an overview of the purpose and progress made in these fields and how they can help improve our quality of life, prevent and treat diseases, and better predict epidemics to reduce mass infections..The use of AI in health and genomic science will revolutionize our future although it will be decades before the majority of population can benefit from it. We don’t have much genomic data yet, and the problem facing scientists today is getting data collection methods to scale. (A quick survey of the audience found that only four out of nearly 50 people had had their genomic sequencing done). At the same time, we are reaching our limits of data processing as humans need AI to process information for us.
While each panelist came to the subject from different angles, a main theme of discussion was the necessity to include social aspects when considering digital health and genetic testing. The smartphone is ubiquitous in today’s world and can be a vital tool to collect data, for example. It gets people much more involved in their healthcare and is moving us away from a doctor-centric healthcare system to a patient-centric system.
Trust was also cited as a key issue. While there is no way you can ever keep people from stealing your data, we can create legislation to prevent them from using it against you, said Dr. Geissbuhler. Ultimately, it is our duty as citizens to share our data (like voting) in order to continually generate knowledge, said Mr. Dermitzakis.
Artificial intelligence allows us to analyze data in ways we have never before been able to. The panel agreed that is was not a predictor tool, but rather a tool for risk analysis. Until now, we were not able to detect or give much attention to rare diseases due to lack of knowledge and resources. With AI, we can assess risk at a much granular level for each individual. If more people share their data there will be less discrimination, as we are all in a risk group, according to Mr. Dermitzakis.
A question from the audience concerned China: Should we be worried about their advances in technology? Mr. Dermitzakis said that China still has outdated systems and the discoveries weren’t there. Iceland has done more in genetics than China, he added. China being such a big market, they are still looking inward. However, it is important to watch what they are doing, added Dr. Geissbuhler. Mr. Salathé rounded off the China discussion with a more generalized statement: “People are still dismissing technology as a technicality.”
Geneva as a center for research and innovation was also discussed by the panel. Opinions varied on the subject: while Mr. Salathé felt that Geneva was lagging behind in terms of research and investment, Dr. Geissbuhler was more optimistic. “Geneva still has a card to play”, he said, citing the recent Swiss Digital Initiative launched earlier this month in Geneva.
To round up the discussion, a member of the audience raised the question most of us wonder when thinking of AI: when will machines take over from humans? All three panelists were reassuring. For the moment, machine learning is concerned with correlation of data. Machines have not yet learned causal reasoning, and we are a priori far from the day they will learn this. The evolution and integration of AI in our lives will be incremental, according to the panel. Scientists and doctors will integrate the technology into their practices, rather than being replaced by it. “It’s not humans versus AI, it’s humans versus humans using AI”, Mr. Salathé stated.
Nonetheless, the panel discussion served as a wake-up call for many to pay more attention to their data, how it is being used and who to trust with it.
Paige Holt, Geneva, September 18, 2019