Event Report: Understanding the Human Brain – Re-thinking Man and Machine

    28 May 2018


    Report on the “Understanding the Human Brain – Re-thinking Man and Machine” discussion at Campus Biotech, Geneva, May 3, 2018.

    About 100 guests gathered in Geneva on May 3 to hear a broad discussion on how a greater understanding of the brain might impact society through technological advances.

    The event was organised by the Swiss Swedish Chamber of Commerce in conjunction with the Human Brain Project (HBP) and the Switzerland Alumni Chapter of Uppsala University. Titled “Understanding the Human Brain – Re-thinking Man and Machine” the evening featured three speakers from within the HBP and two from Uppsala University in Sweden.

    Subjects discussed ranged from biological and artificial intelligence, robotics, brain medicine, self-driving cars, and ethics. Each speaker gave a short talk on their area of expertise before an hour long panel discussion, moderated by Uppsala University’s Dr. Lina Emilsson (senior lecturer at Department of Engineering Sciences, Industrial Engineering & Management).

    <iframe allowfullscreen="true" frameborder="0" height="270" src="https://www.youtube.com/embed/eyHTpQ-cFFw" style="max-width:100%;" width="480"></iframe>

    Edited highlights of the panel discussion (21 mins).

    HBP executive director Christoph Ebell said that while the HBP was not building artificial intelligence (AI) it was attempting to understand biological intelligence and that this would prove useful in being able to assess and decode artificial intelligence for the public and add to informed discussions about its use.

    “If we are deploying AI systems or trying to bring it to market we cannot make the mistake that happened with genetically modified organisms where this conversation with society did not happen at the level it needed to and consequently it was largely rejected. So we are in a situation where we have to seek that conversation and have to be able to help make AI transparent.”

    Dr. Marc-Oliver Gewaltig deputy leader of the HBP’s neurorobotics platform, said most people had an exaggerated view of what AI can do today.

    “For example if you talk about self-driving cars the term itself is a misnomer, these cars are remote controlled, they rely on a huge infrastructure built around them and if this is turned off they will simply not drive.”

    Dr. Philippe Ryvlin, leader of the HBP’s Medical Informatics Platform (MIP), spoke of his hope that the MIP, powered by AI and deep learning, would generate new understandings and treatments for brain diseases.

    “There is already a hospital in the US where every single patent with cancer is submitted to IBM Watson to take a decision on the most appropriate treatment. This will happen in medicine and in medicine of brain disease in the future.”

    “You can imagine yourself that if you are facing a very challenging disease and people offer you to submit your medical file to a system that has proved it will increase your chance of being cured by 30 or 40% you will not hesitate for one second. Of course provided the data will not be transmitted to the public, google or your medical insurers.”

    However, he also warned that the fruits of the MIP would not necessarily happen quickly, and pointed out how few recent breakthroughs there have been in the treatment and management of brain diseases.

    Ginevra Casttellano, Senior Lecturer in Intelligent Interactive Systems at Uppsala University’s Social Robotics Lab, spoke about the importance of social robots, those deployed to assist in classroom or rest homes, being tailored to the needs of those using them.

    She also questioned what was the right level of bonding between users and robots. “Is it always good that people bond, establish a connection with robots, or are their cases when this may trigger some negative effects?”

    “There are also issues about control, should a robot always be in charge? When should it hand over control to a human? How do we decide? Maybe we need some machine learning to help us. Privacy is also widely discussed today, robots will eventually access information about us, so who owns this information and how do we go about assigning responsibility when something goes wrong?”

    Castellano pointed to the recent first fatal crash involving an autonomous car, asking who should take responsibility.

    <iframe allowfullscreen="true" frameborder="0" height="270" src="https://www.youtube.com/embed/wyDK3vkRC_I" style="max-width:100%;" width="480"></iframe>

    Video of the short talks presented by the five panellists (36 mins).

    Dr. Thomas Schön, professor of Automatic Control at Uppsala’s Department of Information Technology, raised the point that out at that people kill each other in car accidents every day in Europe and this is largely accepted because driving is so useful to our lives.

    “I am utterly convinced that in 30 years we will look back and think about what we are doing now, sitting in a junk of steel that weighs about a tonne and which goes about at 100km/h just a couple of metres from each other. And we are relying on our limited capabilities of sensing such a complicated environment. And we have seen that keeps failing all the time and there I see a clear possibility for using this [new technologies].”

    One hundred people attended the sold out seminar held at the Campus Biotech in Geneva. You can watch a livestream of the entire event on Facebook here.