“Generative AI: Pathways to Democratization, Transparency and Sustainability“: Geman-French-Japanese AI Conference in Tokyo

© DWIH Tokyo

On November 12-13, 2024 the German Centre for Research and Innovation Tokyo (DWIH Tokyo), together with the Embassy of France in Japan and AI Japan R&D Network, welcomed 174 participants on Day 1 as well as 74 participants on Day 2 to the Trilateral AI Conference “Generative AI: Pathways to Democratization, Transparency and Sustainability”. This year’s event was a follow-up to three successful AI Symposia in 2018, 2020 and 2022.

The conference opened with remarks from German Ambassador Petra Sigmund, who highlighted the significant economic and social potential of Generative AI, while acknowledging its risks if misused. French Ambassador Philippe Setton emphasized the rapid evolution of GenAI since the previous symposium in 2022 and underscored the importance of ongoing dialogue and collaboration among trusted international partners such as Germany, France, and Japan. Dr. Hiroaki Kitano concluded the opening session by stressing the need for sustainable and universally beneficial development of GenAI.

Keynote: “Democratization of GenAI”

The first plenary session, “Democratization of GenAI,” chaired by Dr. Yasuhiro Katagiri, explored ways to make Generative AI more inclusive and accessible. Speakers addressed the technical and ethical challenges of GenAI, discussing the perspectives of technology developers, users, and organizations. The session highlighted the need for careful consideration of GenAI’s limitations as part of its democratization, ensuring that it benefits the broader public.

Live Stream Recordings:

Greetings/Keynotes “Democratization of GenAI” (Nov 12): YouTube

Presentations:

Keynotes: “Democratization of GenAI” (Nov 12)
Dr. Yasuhiro KATAGIRI (Chair) / Prof. Dr. Judith SIMON / Prof. Céline HUDELOT / Prof. Arisa EMA

Keynote: “Transparency of GenAI”

After lunch, the event continued with a second plenary session “Transparency of GenAI”, chaired by Dr. Anton Zimmermann. The session explored the challenges of bias, hallucinations, and the opacity of decision-making processes in Generative AI. Speakers discussed how to enhance the trustworthiness and accuracy of GenAI and considered who should be held accountable for its harmful impacts and misuse.

Live Stream Recordings:

Keynotes “Transparency of GenAI” (Nov 12): YouTube

Presentations:

Keynotes: “Transparency of GenAI” (Nov 12)
Dr. Anton ZIMMERMANN (Chair) / Prof. Dr. Andreas DENGEL / Prof. Florence D’ALCHÉ-BUC / Dr. Yasuhiro KATAGIRI

Parallel Workshop

Three parallel workshops allowed participants to gain further insight into Generative AI in the industry, technology and research. In “Industry 4.0 / Smart Manufacturing and GenAI”, chaired by Dr. Hiroyuki Sawada, speakers presented real-world applications and case studies of Generative AI applications in manufacturing and industry. The discussion aimed to answer how AI can be effectively used to enhance efficiency in the industry. The workshop “GenAI and the Future of Research”, chaired by Prof. Josephine Galipon, explored the intersection of artificial intelligence, material sciences, and humanistic study. Generative AI is continuously evolving previously existing fields and unlocking new possibilities and innovations, reshaping the future of AI’s role in academic and scientific advances. Therefore, it is even more important to educate society on GenAI’s capabilities and influences. The workshop “Embodied AI and Human-Robot Interaction”, chaired by Prof. Dr. Andreas Dengel, examined how GenAI is being integrated into robotics to facilitate human-robot collaboration.

Presentations

Parallel Workshop: Industry 4.0 / Smart Manufacturing and GenAI (Nov 12)
Dr. Hiroyuki SAWADA (Chair) / Dr. Michael SCHRAPP / Dr. Koichi TAKAHASHI / Guillaume GERONDEAU

Parallel Workshop: GenAI and the Future of Research (Nov 12)
Prof. Josephine GALIPON (Chair) / Dr. Ana ILIEVSKA / Dr. Jean-Claude CRIVELLO / Dr. Kentaro TORISAWA

Parallel Workshop: Embodied AI and Human-Robot Interaction (Nov 12)
Prof. Dr. Andreas DENGEL (Chair) / Prof. Dr. Maren BENNEWITZ / Dr. Jean-Baptiste MOURET / Prof. Tatsuya HARADA

Poster Presentations

The first day concluded with a dedicated session for poster presentations, allowing participants to engage directly with speakers adapting GenAI for diverse applications in their respective areas, such as the medical, creative, research and development field. The evening reception offered a valuable networking opportunity for attendees from Germany, France, and Japan to exchange ideas and reflect on the day’s discussions.

Keynote: “Sustainability of GenAI”

With climate change becoming a pressing matter globally, it is imperative to address Generative AI’s sustainability. In the second day’s plenary session “Sustainability of GenAI”, chaired by Dr. Florence Ho, speakers discussed strategies for reducing GenAI’s environmental impact, emphasizing the importance of sustainable AI development. At the same time, they explored the potential of GenAI to contribute to broader sustainability goals.

Live Stream Recordings:

Keynotes “Sustainable GenAI” (Nov 13): YouTube

Presentations:

Keynotes: “Sustainable GenAI” (Nov 13)
Dr. Florence HO (Chair) / Johannes Leon KIRNBERGER / Dr. Marc DURANTON / Prof. Rio YOKOTA

Roundtable Discussions for Future Collaboration

The final session of the trilateral AI conference featured nine roundtable discussions. Upon realizing similarities and shared interests of a few individual groups, some joined together to collaborate and combine their perspectives and approaches. The roundtables allowed a concentrated discussion in a smaller and proactive setting.

Recap Session

During the recap session, moderated by Sabine Schenk (Heidelberg University Office Kyoto), representatives from each roundtable summarized their discussions, presented their findings, and shared key insights. Dr. Jean-Baptiste Bordes (Embassy of France) provided a comprehensive summary of the entire conference. The event concluded with closing remarks and expressions of gratitude from Axel Karpenstein (DWIH Tokyo), Dr. Jean-Baptiste Bordes, and Takanori Murakami (AI Japan R&D Network), acknowledging the collaborative efforts of the three nations that made this stimulating conference possible.

The Importance of Continued Collaboration

Since the last AI symposium in 2022, the field of Generative AI has seen significant advancements, particularly with the rise of tools like ChatGPT. Over the past two years, GenAI has become more widely accessible, with increasing use across personal, academic, and business domains. As software providers continue to integrate GenAI into their services, the technology’s ability to generate text, video, and speech has rapidly improved. This year’s conference underscored the need for continuous evaluation and international collaboration to ensure that Generative AI develops in a way that aligns with shared democratic and ethical values.

Report: Sophie Braun (DWIH Tokyo intern)

Updated on: December 3, 2024