AI positioning: data governance and effective use are needed
This year's general meeting of the Swiss Data Alliance focused on artificial intelligence (AI). Ruth Fulterer's assessment of the current situation was followed by a panel discussion on the applicability of AI for companies.
Ruth Fulterer, editor in the Science, Technology and Mobility section of the Neue Zürcher Zeitung, addressed the reality and promises of artificial intelligence (AI) in her assessment of the current situation. Whether the construction of data centers for AI could actually quadruple a country's gross domestic product (as Keith Strier claimed at this year's WEF) is doubtful. However, the enormous technological change of recent decades cannot be denied. This can also be seen in everyday journalistic work. Although AI is also used at the NZZ - for example in the formulation of search engine-friendly titles - the question of the increase in productivity and the effective benefit of AI is very difficult to quantify. The answer varies greatly depending on the application.
"Arms race" as a political argument
Fulterer also questioned the narrative of the "arms race" in the development of AI. Against the backdrop of the copyright debate surrounding the use of copyrighted training data, the argument bordered on "rhetorical blackmail". It should also not be forgotten that a lot of state money is at stake, as well as access to energy resources and economic interests, for example in the production of chips. "Will a very capable model like DeepSeek really determine world politics?" asked Fulterer. And immediately answered himself: "I don't see it yet."
Benefits along the 80-20 principle
Overall, the benefits of AI are probably along the lines of the 80-20 principle - in other words, a lot is possible, but the last few percent of the work is still very time-consuming. Technology companies sometimes like to play down these outstanding development steps. This was also expressed at a Tesla party, where robots were used as bartenders and impressed visitors with their skills. As it turned out, it was people wearing VR goggles and motion capture suits who were controlling the robots.
The panel of Ruth Fulterer, Marc Holitscher and Alain Gut addressed the challenges and opportunities that AI offers for business and society. They spoke of the need to strategically anchor AI in companies, with the use of data being particularly crucial. Companies need to understand how they can use AI models efficiently in order to create real added value.
Data quality is crucial
Data quality is a key issue here. Without high-quality data, most AI applications are not useful. Many companies have not invested enough in data strategies in the past, which presents them with the challenge of fully utilizing AI today. In addition, the use of AI is not only a technical challenge, but also a cultural one. The transition to a data-driven company often requires far-reaching changes in the way employees think.
Finding niches
The opportunity for Switzerland is to adapt technologies quickly, even if it does not always develop them itself. Highly specialized small and medium-sized companies could play a pioneering role in niches. However, many companies have not yet found the right use cases for AI and show a certain amount of uncertainty as to how they should use the technologies sensibly.
Promoting digital skills
Last but not least, dealing with AI requires not only technical skills but also a deeper understanding of the ethical and social implications. A complicating factor is that digital skills in Switzerland are less developed across the entire population than in the Nordic countries, for example. It would help to consistently focus on "digital only", because only then can accessibility and usability be guaranteed for the entire population.
