by Peter Kunz (ERCIM Office)
ERCIM and the Japan Science and Technology Agency (JST) held their second joint symposium on 8 and 9 December 2021. The symposium aimed to present future visions and recent research results conducted in the frame of Japanese Advanced Integrated Intelligence Platform the (JST AIP) project as well as from European institutions. Some 50 scientists participated in this event to share their recent research and identify collaboration opportunities in the context of the European Horizon Europe framework program or relevant initiatives from JST.
The joint workshop focused on the theme “Accelerating digital transformation with trust for a post-COVID-19 society”. It presented recent results, emerging design frameworks and technical solutions for dealing with the coming social changes in the era of digital transformation. In a remote and contactless environment, sophisticated trust concepts and technologies are strongly demanding in every social and technology domain. Furthermore, it also provided an opportunity to share the different situations in Japan and Europe concerning the use of big data for analysis, the protection and preservation of privacy, the adherence to principles of human centric artificial intelligence (AI), and several standardisation efforts. Of particular interest were results from applications of AI in areas such as medical diagnosis, bioinformatics and drug discovery. In addition, the symposium covered a wide range of topics of common interest within the broader AI, Internet of Things (IoT) and big data areas.
The symposium was structured in four themes: (i) Trustworthy AI: theory and systems, (ii) Mathematical approaches to privacy and security, (iii) The future of IoT and AI, and (iv) Digital governance, ethical, legal and societal impacts, and AI.
The first day started with a keynote talk by Prof. Takayuki Ito (Kyoto University) entitled “An agent that facilitates crowd discussion”. He presented a large-scale online discussion platform called D-Agree. Such platforms require support functions that can efficiently achieve a consensus, reasonably integrate ideas, and discourage flaming.
Two parallel sessions then investigated the topics “Trustworthy AI: theory and systems” and “Mathematical approaches to privacy and security” In the Trustworthy AI session, Prof. Isao Echizen (National Institute of Informatics) presented "Real or fake? From biometric data protection to fake media detection", and Tim Baarslag (CWI) spoke about “Coordination of intelligent and autonomous systems through negotiation”. In the parallel session “Mathematical approach to privacy and security” Jun Sakuma (Tsukuba University) gave a presentation “Towards the realisation of AI trusted by humans” followed by Michele Sebag (CNRS – French National Centre for Scientific Research) who spoke about “Extremely private supervised learning”.
Then, all participants came together in a panel session with the speakers to discuss questions about the relationship between their work and trust, the problems to be solved by AI and how can they be solved.
Prof. Fabio Martinelli (CNR IIT) opened the second day with a keynote on “Data usage control for data sovereignty”. He introduced the notions of data centric policies, policy refinement and policy enforcement in several scenarios, including cyber-threat intelligence management.
The symposium continued with two parallel sessions. “The future of IoT and AI” was the title of session C, where Prof. Takayuki Nishio (Tokyo Institute of Technology) gave a presentation on “Distributed machine learning in IoT networks” followed by Helmut Leopold (AIT – Austrian Institute of Technology) who spoke about “IoT and AI for a sustainable digital future”. In the parallel session entitled “Digital governance, ELSI, and AI”, Prof. Minao Kukita (Nagoya University) gave a presentation “AI is the message: How AI can affect our view of the humans” and Prof. Alexander Schatten (SBA Research) discussed the question “Is the extended mind embracing artificial intelligence? Ethical and social consequences and responsibilities.”
In the following panel session, the speakers of the parallel sessions summarised the role of the researchers in the development of autonomous algorithms and the impact of AI on the society. The speakers and the organisers ended the symposium with the desire to strengthen the research network and to establish further cooperation between groups.
Dimitris Plexousakis, ICS-FORTH, Greece