Finally Taking AI Seriously on the Political Agenda
Events#ZEWlive on the EFI Report 2024
Professor Irene Bertschek, head of the “Digital Economy” Unit at ZEW Mannheim and deputy chair of the Commission of Experts for Research and Innovation (EFI), presented the EFI Report 2024 in a new edition of the digital series #ZEWlive on 12 March 2024. The focus was on the topic of artificial intelligence.
After her presentation, she discussed the key findings with Dr. Andreas Liebl. He provided insights from his perspective as founder and managing director of appliedAI Initiative GmbH in Munich, which aims to act as a kind of personal trainer to help companies get fit for AI.
EFI Report 2024: AI is a key technology
Bertschek began by outlining core findings of the EFI report in terms of AI. As a key technology, AI is highly relevant for the transformation of the economy and holds significant potential for innovation and growth. China and the US dominate technology development, while Germany and the EU27 lag behind. Since there are no major tech companies in Europe, there is a need for an AI ecosystem with contributions from academia, small and medium-sized enterprises (SMEs), and start-ups. Basic research in Germany should be vigorously supported.
One of the main problems is that the infrastructure in Germany and Europe is inadequate. The computing capacity available in Europe trails behind that in the US. There is therefore a need for public-private cooperation to invest in better infrastructure. In addition, a specific obstacle for Germany is the strict interpretation of data protection, even compared to other European countries subject to the same General Data Protection Regulation. Germany must therefore consistently implement its data strategy, as AI models are trained and applied on the basis of data.
It is also crucial to build and expand competencies for AI – not only for the development of AI but also for handling this technology. Germany’s agreement to the EU’s AI Act is the right step. However, due to the dynamic technological development, there must be room for regulatory learning to adapt the AI Act over time.
AI infrastructure must meet business needs
According to Liebl, there are large-scale AI applications in companies in China. The technology is much more widespread there than in Germany. In the Chinese economy, AI is used much more professionally. However, China is not more innovative than Germany when it comes to AI. The clear leaders in AI technology are the USA, with a well-established tech scene and strategically positioned start-ups.
Liebl emphasised that the public datacentres currently available in Germany, which are mostly operated by research institutions, do not always meet the requirements of AI companies. If there is a strategic expansion of public infrastructure, it must consider the needs of businesses.
Policymakers must finally take AI seriously
Liebl called for AI to be recognised as a key technology and to be taken seriously at the political level. Overall, there is a lack of politically defined success indicators, known as output key performance indicators. Just as no company can function without electricity and the internet today, AI will become a fundamental requirement for competitiveness. Everyone needs to engage with AI, starting with school and university education, especially considering that deep fakes could have a strong impact in a European election year like 2024. However, AI should not be prioritised over other political goals but should be seen as a means to address other significant societal challenges, such as climate change or the energy transition.
