Up until now, European companies have been holding back when it comes to using artificial intelligence. According to a recent survey by Bitkom, Germany's official digital association, only 15% of the companies surveyed in Germany use AI. That despite a full 73% being aware that the technology would be groundbreaking for them. On our quest to learn the causes, we interviewed three experts from the industry during AI DAY. Karina Buschsieweke, CEO of Lana Labs, sees the reason for the disparity in the fact that German medium-sized companies have long been able to operate successfully without the need to go digital. The industry has been doing well for a long time in Germany, adds CEO of Lengoo and member of the board at Bitkom, Christopher Kränzler, and it is the reason why people have only ever tried to improve incrementally rather than driving change. This way of thinking has put Germany in a tight spot, because even during a crisis when resources have to be saved, the possibilities of digitalization are certainly not the first thing that comes to mind. And that despite the fact that financial pressure is growing exponentially in the wake of the crisis. The only way of getting out of this situation is to save resources, which usually leave just two options: laying off staff or selectively reducing the use of resources in other areas. If companies are, however, not aware of the latter option, the consequences for staff can be drastic—a situation that benefits nobody in the end since companies with reduced team sizes also suffer from a reduction in productivity and innovation—not to mention the loss of trust among the remaining staff and the consequences for those who have lost their jobs. According to the latest Deloitte AI study, the use of artificial intelligence in Germany pays off faster than in other European countries. In most companies, AI starts to pay off financially in under two years, while 42% even put the figure at just one year. As such, the use of this technology offers the unique opportunity to reduce costs quickly, but also in the long term.
Skepticism and uncertainty in handling data
This hesitation stems from the huge amount of uncertainty when it comes to handling data. Not many know, for example, what exactly the GDPR says and what it means for the handling of data. Dr Thomas Ramge, Research Fellow at the Weizenbaum Institute and author of several books on the topics of digitalization, data, and changes in the world of work, speaks of a “deep-seated skepticism that anything to do with large amounts of data tends to pose a risk.” This makes us lose sight of the fact that something intelligent—a basis for making better decisions—can be created from this wealth of data. “This is a huge problem for us,” says Ramge. Christopher Kränzler points out that this skepticism affects our private lives just as much in Germany. He speaks of a “general, slow technology adaptation” in the country and cites as an example that not one person in his circle of friends owned an iPhone before the release of the fourth generation. “This is a sign of the mentality that makes people simply observe things for far too long before actually taking action themselves.”
Dr Ramge considers our culture to be the problem as we are strongly influenced by a fear that data could be misused or that we might become “dependent on some superstar company that knows more about us than we know about ourselves.” This image, he explains, has been further fed by films and books, and so it comes as no surprise that “the initial reaction is negative or at least dismissive, even if no personal data is involved at all.” But this fear is not without consequences: it leads to a tendency to when in doubt, make savings in the wrong places.
The myth that cracking down on data protection would create a competitive advantage
One assumption that receives particularly harsh criticism is that the German approach to handling data (the German "Datenschutz") means a competitive advantage for Germany because everyone who uses our systems trusts them. According to Dr Ramge, this myth has worked neither for Cloud Computing nor for Big Data, which is why it will not work for AI either. Instead, Europe needs to understand that non-personal and non-trade secret-related data has to be shared to create a data space in Europe that is conducive to innovation. After all, according to Dr Ramge, “data is the most important resource for innovation in the 21st century”. In addition, Europe already has an excellent research landscape and enough talent—but still no open data solution for the European area. He sums up: “In terms of AI, our attitude towards data is very clearly something that is culturally paralyzing us to such an extent that it is significantly hindering the further development of data-rich systems.”
Understanding the crisis as an opportunity
On one point everybody agrees: there is no distrust in companies, but “rather the fear that something will end up going wrong,” says Karina Buschsieweke. But this is the case with any new project and with any new technology—that is to say, entirely independently of the topic of AI, emphasizes Buschsieweke. Christopher Kränzler points out that trust is undoubtedly not an issue at Lengoo. When training engines for machine translation, the most specific data possible is needed to create company-specific engines, as such, there is no point in mixing data from different companies, and customers know from the start that there is no need to worry. The problem is the companies’ lack of knowledge when it comes to the advantages of AI solutions. This is why, according to Christopher Kränzler, the creation of knowledge is fundamental so that companies understand what a “treasure trove of data they possess—a treasure for which American companies have long since been buying out entire companies.” This is why it is essential to be done with myths because, according to Lengoo's CEO, “it is not about automating everything with machine learning, but rather finding a small lighthouse project that you can learn with and show all the other players in the ecosystem what advantages artificial intelligence ultimately has.” Kränzler wants to see the advantages of AI applications also being worked out at a political level, though he points out that we are “running out of time.” Now, he continues, is an excellent time to “wake up in this crisis,” and companies need to finally understand that the crisis offers the opportunity to trigger a complete transformation towards AI. Karina Buschsieweke agrees that crises, in particular, can provide food for thought, cause the status quo to be questioned, and improvements to be actively sought. After all, in times of crisis, the use of AI creates much-needed transparency, though this is also necessary with clients. The CEO of Lana Labs explains that an honest exchange with clients is what creates the most trust. “What kind of data will we include in the analysis? Does it include personal data? How will it be pseudonymized or not evaluated at all?”—all these questions need to be answered.
Instead of laying off staff when financial difficulties arise in times of crisis, companies should consider where they can really make long-term, sensible savings. The most significant and most long-term potential for saving is not to be found in the area of personnel, but rather in processes that are currently inefficient, which Jonathan Wuermeling, host of the AI DAY and Head of Brand at Lengoo points out at the end of the panel discussion by quoting the 15 trillion dollars mentioned in Karina Buschsieweke's keynote speech, that are lost every year as a result of inefficiencies. “Perhaps we should not always look just at cost optimization,” suggests Buschsieweke. After all, process improvements are worth just as much as money saved, since that is precisely what they amount to in the end.
Watch the whole panel discussion on the topic on Youtube (German Only):