The race for artificial intelligence (AI) supremacy is heating up on a global scale, with Europe striving to keep pace with the likes of the United States and China. However, a recent conference organized by the European Court of Auditors (ECA) shed light on some of the key challenges Europe faces in its AI development journey.
Data Center Deficits Hinder AI Growth
One of the most pressing issues highlighted at the conference is the lack of sufficient data centers in Europe. Audrey Herblin-Stoop, Head of Public Affairs at French AI company Mistral AI, emphasized the critical role data centers play in training large language models (LLMs). These models, like Mistral’s own “Mistral Large,” require massive amounts of data and processing power to function effectively.
“We are reaching the capacity; we need to build datacenters and ensure there is enough electricity for the scale of AI development today,” Herblin-Stoop stated. “The access to infrastructure is important, if you don’t have this, you cannot build large language models.”
Mistral AI itself benefited from access to supercomputer facilities provided by the European Commission to train its models. However, Herblin-Stoop acknowledged that such access might not be readily available for other companies in the future, highlighting the need for significant infrastructure investment.
The Financial Firepower Gap
The ECA’s report further emphasizes the financial disparity between European and American AI companies. Killian Gross, responsible for AI policy coordination at the European Commission, readily admitted that “EU companies lack the financial fire power” compared to US giants like Microsoft. This lack of resources makes it difficult for Europe to compete on a level playing field.
Gross, however, remained optimistic, stating, “We cannot catch up that quickly, but this is not a reason to be frustrated – it’s a reason to strengthen our efforts.”
A Patchwork of National Strategies
Another hurdle Europe needs to overcome is the lack of a unified approach to AI development among its member states. MEP Dragoș Tudorache, who played a key role in shaping the European Parliament’s position on the AI Act, pointed out the discrepancies in national AI strategies across the continent.
“Enabling the take up of AI is something that member states have to put in place. But national strategies see a very scattered landscape. And the contents of the strategies are not all the same,” Tudorache remarked.
This lack of harmonization could hinder Europe’s overall progress in AI development.
The Data Dilemma: Access and Diversity
Data access is another crucial factor for European AI companies to flourish. Herblin-Stoop stressed the importance of having access to large amounts of data in various languages to ensure their models are culturally relevant.
“We need to find a way to have access to data in a better way to have other cultures represented,” Herblin-Stoop said. “The content will come from the US and China in the future. But as a global company we do care about building models that are strong in languages other than English.”
This concern highlights a potential bias in AI development, where models trained primarily on data from certain regions or languages might struggle to understand or respond effectively to information from other cultures.
The AI Act: A Step Towards Regulation
Despite these challenges, Europe is taking significant steps towards a responsible and regulated AI future. The signing of the AI Act this week marks a major milestone. This new set of stringent rules aims to govern high-risk AI systems and ensure ethical development and deployment.
The general-purpose AI rules within the Act will come into effect in May 2025, while regulations for high-risk systems will take effect three years later.
The Road Ahead: A Call for Collaboration
The race for AI dominance is far from over. While Europe faces challenges in terms of infrastructure, funding, and data access, the continent is making strides towards responsible AI development through initiatives like the AI Act. Collaboration between governments, research institutions, and private companies will be crucial for Europe to bridge the gap with the US and China and establish itself as a leader in the ethical and responsible development of AI technologies.