On 14 October 2020, EIF organised a debate on ‘Emerging Technologies: AI and Supercomputing’, aimed at taking stock of what is happening at global level and investigate what Europe is doing to secure its place as a leader in this geopolitical tech race.
The discussion, moderated by EIF Director General Maria Rosa Gibellini, was hosted by the MEPs Pilar del Castillo, EIF Chair, and Eva Kaili, EIF Steering Committee Member, and featured the following speakers:
- Lucilla Sioli, Director of Artificial Intelligence & Digital Industry DG CNECT, European Commission
- Dr. Zhou Bin, CTO, Huawei Ascend Computing
- Dr. Alessandro Curioni, Head of IBM Research in Europe, Director of the IBM Research Lab in Zurich
- Dr. Jacques Bughin, CEO, Machaon Advisory; Advisor, Fortino Capital and Venture Partner, Antler; Professor Solvay Business School
Pilar del Castillo MEP opened the floor underlining how the economic potential of AI has never been stronger; for this reason, Europe must take action to ensure that data can flow among EU-wide common and interoperable data spaces. Europe must also promote investments in the next generation tools and infrastructure to store and process data. She stressed the link between AI and computing capacity and how Europe must focus on reinforcing capacities in high-performance computing.
She welcomed the European Council conclusions of the 2nd of October and praised the allocated 20% of the recovery and resilience facility to the digital transition.
Eva Kaili MEP put the attention on how the pandemic accelerated the challenges derived by new technologies and how it is crucial to respond to these issues globally and with decentralized solutions. There is the need to invest both in software and hardware to achieve the maximum capacity and capability to use these technologies, like creating common standards to accelerate some processes.
To develop a holistic strategy on AI and understand how to set ethics by law and by device is a priority. She also stressed the need to set a data governance, in order to create a better ecosystem where competition can work and all players can really be involved. “We need to be able to move really fast, this is a global race and we want to influence and impact it with quality.”
Lucilla Sioli started her intervention stressing how the capabilities of running AI on supercomputers will bring increasing benefits. Computing requirements are very important for the next generation of AI, increasingly based on very large sets of unstructured data. At the same time, the processing of data is very energy consuming so it is crucial to come up with energy efficient approaches.
She reminded how the application of AI into a supercomputer has helped to identify the drugs that could help ill patients during the Coronavirus pandemic.
Ms. Sioli concluded giving a clear overview of the European Commission’s current work, investing in both AI and supercomputers in the European Union, on the research and deployment of these tools, focusing on the regulatory framework and promoting a trustworthy approach to AI.
Dr. Zhou Bin reiterated how AI is helping fighting the Covid-19 pandemic. At the same time, there are a lot of upcoming challenges for the computing architecture. Huawei is currently redesigning the whole processing architecture; all this necessary rebuilding requires a lot of investment but will help the industry, academia and researchers to facilitate their researches. The technology must at the same time be able to protect the privacy and the safety of data.
He also stressed the need to have unified standard for interoperability and collaboration between different sectors. AI, together with supercomputing, will create a new era for computing technologies and industries and, at the same time, shape a better world.
Like other speakers, Dr. Alessandro Curioni mentioned the Covid crisis and how it created a wakeup call to for everyone to find a way to better solve and tackle future challenges. There is the need to create the right ecosystem with a proper data sharing mechanism and develop a proper infrastructure able to empower AI.
All this needs to be embedded into a new concept of community discovery and collaboration because it is not something that can be handled and tackled by a single country or institution. Supercomputing and AI can help boost the scientific method and in order to encourage it, conventional approaches to supercomputing must fuse with new approaches that make AI and algorithm implementation more sustainable and, eventually, also converge with quantum computing. All this, delivered with a platform allowing the maximum interoperability.
Dr. Jacques Bughin underlined the fact that we are at the start of a migration towards a powerful knowledge economy that sees many building blocks (supercomputing, AI, data) still to be put together. He agreed with the need to change the architecture of AI, supercomputing and data. Europe should not miss the opportunity to take a leading role and promote standardisation since, in the current global race where a few countries and firms rule, the battle for control is apparent through a set of various strategies such as user standardisation, or patenting.
According to Dr. Bughin, Europe is not leading but has some powerful plays to be a contender of relevance such as its positions in AI and robotics, its dense network of digital start-ups and its lead in the race of supercomputers.
He finally shared the importance of open data for all the actors part of a level playing and how AI should not be taken without sustainability, an element that differentiates Europe compared to US and China.