AI Stakeholders met in Brussels to Discuss on AI Ecosystem & Strategic Research Agenda
Almost 40 people from the AI research community and policy ecosystem gathered on October 19th, 2022, in Brussels in a 2nd ICT-48 Community Workshop organized by the VISION project.
Representatives from VISION, HumanE-AI-Net, ELISE, TAILOR, AI4Media as well as ELSA, AI4Europe, EuROBIN, Adra-e and the European Commission spent a day discussing how to collaborate with each other, leverage on each other’s results and what could a mapping the AI ecosystem, and a European AI Strategic Research Agenda look like.
The morning part of the programme featured 6 roundtables, where participants could discuss and exchange ideas on topics like:
- cross-network scientific challenges
- educational activities
- collaboration with DIHs and with the industry
- cross-network communication
The European Commission, represented by Cécile Huet, kicked off the afternoon sessions outlining a vision of a vibrant network and ecosystem of AI excellence centres working together as a way to develop Europe’s position in AI (research) and support industry and society with AI application “from the lab to the market”.
In a collective mapping, the group – led by Maurits Butter (TNO) and Iddo Bante (ADRA-e) – discussed
- who are the most important users of such mapping
- how to engage the target audience
- efficient organisation of the collective mapping
Discussions revolved around whether there is a need for such mapping, recognizing the importance of creating a common understanding of what is mapped and what is understood by excellence, and the need to connect any topic with the actual application. The enormity of the topic and effort was recognized and it was decided that further alignment and discussions might be needed.
Another key topic of discussion revolved around the idea of a European AI Strategic Research Agenda that complements the individual SRAs of the AI networks of excellence. Moderated by Fredrik Heintz (TAILOR), the session featured a presentation from each of the networks of excellence on the critical points of their SRAs. The discussion then turned to how these diverse topics and application areas can be connected and draw higher-level conclusions.
A first start to come up with a potential structure connecting the various topics in a categorization (also to initiatives like ADRA SRA) was presented by Freek Bomhof (TNO) with a discussion on how this categorization connects to the research communities and how to introduce dependencies. It was agreed that collaborating on creating an aligned view of a joined SRA can contribute to the analysis of commonalities and differences in the different AI application domains and has the added value of connecting and learning from what the various communities think on certain topics. The group also agreed that creating a common higher-level vision should be carefully phrased to leave room for the networks to position themselves, express positions, and be realistic in scope, detail and ambition. A group of representatives from the networks will further continue the work started.
The overall conclusion of the workshop was that the AI ecosystem is developing and open discussions among the various communities to enhance collaboration and peer learning.
Further, a session in the European Big Data Value Forum EBDVF 2022 in November will aim to collect further insights from the AI community on the presented categorization.