Opinion Article by IE.F Managing Director Clark Parsons in the WELT
WELT.de / Updated: 12.09.2024 / Opinion Article /
© WELT (originally in German)
Image Source: WELT (Originals: Yuichiro Chino/Getty Images; Marc Melkenbeek/Picture Alliance for DLD)
Europe's economy can benefit enormously from its untapped wealth of data, writes guest author and think tank boss Clark Parsons. German and European companies should not let this treasure out of their hands. It is the key to industrial liberation.
Europe's data may be our last hope in AI. When former ECB President Mario Draghi presented his report on European competitiveness on Monday, he mentioned many things that we have seen and accepted for years. Among other things, he pointed out that we in Europe have not developed any global players in the technology world and that a handful of big tech companies from the USA dominate the global market. However, this should no longer come as a surprise to anyone.
Germany and Europe are now lagging far behind the big names from the USA. Unlike European companies, the Apples and Microsofts of this world have vast amounts of user data of all kinds, unprecedented computing power, easy access to billions of customers and biblical amounts of investment capital to consolidate their dominant position. In 2023 alone, two-thirds of all venture capital investment in generative AI companies in the US came from Google, Microsoft and Amazon.
In response, Mr. Draghi rightly argues that Europeans must not only regulate things, but also tackle them in order to harness their innovative power. But in view of the gigantic market power that we now face, many are asking themselves: How can we still hold our own and where should we even start?
Let's start with the AI revolution. Because it still holds enormous opportunities for Europe. And the key to Europe's AI opportunities is our data. Why is our data so important? It is untapped, characterized by incredibly deep and specialized domain knowledge - and it is the fuel for AI.
Data trains AI models. Back in 2000, a Microsoft paper showed that an AI model can deliver better and more accurate results if it is simply fed with more data. Since this became clear, the tech giants have been racing to collect breathtaking amounts of data. You can see this with every new language model that comes out: there is now talk of billions of “tokens”, i.e. data records.
But this race hasn't always been above board lately. The New York Times discovered this year that OpenAI had been training its ChatGPT models with pretty much all YouTube videos - all without permission, quietly and secretly. Why didn't the rights holder Google complain about this? Rumor has it that Google did the same and didn't want to attract attention. A pact of silence about our data.
However, it is a misconception that we have now lost this race because the data treasures of the consumer Internet are now distributed among others. After all, most data is not on the public Internet, but is held by companies and organizations - deeply buried in complex and often hidden value chains. And here comes Europe's opportunity.
Companies around the world are trying to find out how AI can change their business. Whether it's improving processes, developing new products and services, or boosting productivity. Many of these innovations will come from dynamic start-ups that are now entering the market.
A new policy paper from the Innovate Europe Foundation shows that all these start-ups can benefit from this untapped treasure: Our unique domain knowledge - our industry processes, company IP, trade secrets, etc. - can and must be used in a data protection-compliant way and with the help of AI. This will be the key to digital and industrial liberation in Germany and Europe.
Data sovereignty is the only solution
Handing over this data to non-European companies would be a big mistake. Data sovereignty is therefore the only solution. Of course, along the way there will be many temptations to take the easy way out, especially when large US technology companies present seemingly harmless tools that then secretly collect information and get involved in important processes. But we know better: He who holds the data holds the power.
And it's not just companies that can benefit. There are several public sectors in which Europe could shine with AI solutions. Two examples are healthcare and mobility.
In the healthcare sector, our gigantic public healthcare systems and research institutions represent an incredible pool of knowledge and insights for AI-based innovation. There is much work to be done to standardize data collection while ensuring that personal data, privacy and secrets are protected.
But the benefits to society are immense and the potential for European companies to be at the forefront of this innovation movement is enormous.
Moreover, nowhere else in the world is there such a concentration of public transportation and multimodal mobility networks as in Europe. The potential here is also huge, but the temptation to simply let Big Tech from the USA take control is just as great. For example, one could question why local and national rail and rail systems first share their real-time data with Google Maps before passing it on to local mobility platforms such as Omio in Berlin.
Both Europe's SME champions and its public institutions have data and specialized domain knowledge that can usher in a new era of innovation in collaboration with start-ups. It is extremely important that we do not misunderstand data sovereignty. Data sovereignty does not mean being stingy with data. It doesn't help us at all if we don't do anything with it for fear of data loss.
More security could come from an “infrastructure made in Europe”. Without our own infrastructure, from cloud providers and data centers to our own chip production, we will always be dependent. Fortunately, Germany and Europe have many players of their own that offer cloud-based services ranging from storage to office utilities.
It would therefore be a strong signal if our administrations were to make more use of them. A flagship project here is Baden-Württemberg's AI-based F13 platform, a public-private partnership for greater efficiency in administrative procedures and processes, which is now available nationwide.
While the hype surrounding large AI language models is already dying down, the services and applications that are built on them will permanently change the way we live, communicate, learn and do business. So the window of opportunity has not closed yet. That's why we need to listen to Mario Draghi and set the course so that our efforts to create an independent innovation and AI ecosystem merge with our industrial expertise.
Data sovereignty should be our cornerstone for AI and the bold use of data for AI must be the result. Then Europe will have a chance to play at the top again.
This article was translated by the IE.F for the purpose of this blog post only