This post was inspired by a recent publication by the European Business Magazine.
It is no secret that the European Union is interested in homegrown artificial intelligence technology, which is no surprise given the shifting geopolitical landscape on the international scene (trade war, anyone?) combined with the transformative impact of AI on nearly every industry. No one knows exactly how AI adoption will shake out in the long run or what that world will look like exactly, but the surface-level tangible benefits to productivity are so clear that Europe cannot risk being fully reliant on trade partners to have access to the technology. Moreover, in light of the demographic shift in the Union’s workforce, such productivity gains are essential for offsetting the fiscal strains imposed by degrading dependency ratios. What is a continent to do?
In light of geopolitical realities, the tone in Brussels towards AI has quickly shifted gears this year. Less than a year ago, the EU was proudly celebrating its protection of consumers and civil liberties with the adoption of the EU AI Act, which designed a regulatory framework for how general-purpose or (in the case of sensitive use cases) application-specific AI models are to be deployed (or not) onto the European market. Now, this year, the European Commission’s north star is the word “competitiveness.” Look no further than President von der Leyen’s competitiveness compass, announced this January.
So, where does this leave AI in the eyes of European governments? In a bit of an awkward spot, to be honest. On the one hand, the hard work of defining the implementation of the AI Act is now underway, but many industry representatives believe such measures should exercise restraint as much as possible so as to not undermine Europe’s ability to develop AI technology themselves. This debate is ongoing, and industry practitioners from both large enterprises and startups will need to follow this process closely as it unfolds.
Policy aside, how can we define Europe’s “hierarchy of needs” with respect to AI? In broad strokes, we can define two levels (in descending order of importance):
- The need to have reliable access to AI technology
- The ability to create novel AI technology themselves
If Europe is unable to achieve #1, they will be unable to compete on the world stage as competitors around the world benefit from the productivity gains yielded by AI.
If they cannot achieve #2, they will be still be able to benefit from the gains yielded by #1, but these will be bottlenecked by (a) the availability of open-source software which is suitable for each of the 24 official languages of the Union, (b) commercial terms set by foreign trading partners (which will represent an unrecoverable input cost to most business processes), (c) trade policy decisions made by those partners’ governments (export controls, etc.). As such, you can see why the EU views both of these as imperative.
So, what is Europe doing to achieve the AI gains they are looking for? In large part, they are ramping up public spending on compute resources. As scale-up funding is often a bottleneck for startups trying to stay in Europe, the Union hopes that so-called “gigafactories,” at a cost of an estimated €20B will enable startups such as Mistral AI and researchers to keep up with American and Chinese big tech spending on AI development.
With years of experience designing and developing AI systems and advising on AI regulatory developments, Nodora Partners is positioned to help you navigate this evolving landscape. If you or your team would benefit from expert guidance, do not hesitate to contact Nodora Partners for a free consultation today.