Researchers as well as industry associations welcome the fact that the German government has recognized the importance of artificial intelligence (AI) and recently launched a three billion dollar AI strategy. However, they insist on keeping an eye on implementation in practice, especially in small and medium-sized enterprises.
“It is good thing that the Federal Government has agreed on an AI strategy with us at the Hasso Plattner Institute (HPI) and is also investing additional funds for this," HPI Director Professor Christoph Meinel stated. “Because the three billion will help our innovation site considerably,” Dr. Johannes Winter confirmed, Head of the Office of the Learning Systems Platform in Munich. This is especially so since the three billion will have leverage effect. It is estimated that joint projects with industrial partners and support initiatives by the federal states will bring the total to approx. EUR 6 billion. Dr. Winter: “No other European country is investing that much.”
In addition, the German research landscape is already very well positioned, Dr. Winter stated. “This applies to basic research, but above all to application-related research as well. We are a global leader in Industry 4.0, autonomous driving, AI-controlled prostheses or real-time analytics against credit card fraud.”
However, VDMA warned that neither the budget of three billion euros nor the announced 100 new professors would guarantee the successful use of artificial intelligence. “We don't need research in an ivory tower, but in the end we need intelligent solutions on how to use AI in companies as quickly as possible,” Hartmut Rauen said, Deputy Chief Executive Officer of VDMA. “Unless we succeed in transferring technology from basic research to industrial practice, artificial intelligence will remain a topic for the proverbial scientific ivory tower.”
HPI Director Meinel also believes that the greatest challenge is to transfer the good research results of the internationally recognized German AI research landscape to the economy in such a way that the value added there remains at the business location. “This was too often not the case in the past.” Building AI expertise in strong German fields such as mechanical engineering is not a small task, Mr. Meinel said. “The reason for this is that our location is characterized by strong and diverse medium-sized companies that simply cannot afford the huge investments in research and development that Alibaba or Amazon are making right now.”
The VDMA therefore calls for targeted support of SMEs. “AI can only become a real success story if it succeeds in bringing technology to a wide range of industrial SMEs.” Consequently, ensuring efficient technology transfer is important. “The AI strategy of the Federal Government provides that AI trainers from the existing ‘SME’ competence centers go to the companies and support them in examining their workflows and business models for potential AI applications,” Mr. Meinel reported. “That's at least a start and a good approach.”
However, a lack of awareness still exists in many companies. For example, Mr. Meinel pointed out a recent study commissioned by the Ministry of Economic Affairs, according to which only five percent of German companies already use AI although a quarter of them are involved in artificial intelligence. “In the study, a total of three-quarters of the companies surveyed said that AI is not relevant to them. I think this is a mistake in many cases,” the HPI expert stated.
Dr. Winter sees another challenge in the fact that Germany can only succeed in the international race for the top position in artificial intelligence together with Europe. “We must not consider Germany as a single player, but as one of the performers in the European team.” The creation of a European data area, as envisaged in the AI strategy, would therefore be the right way forward.
At the same time, VDMA warns politicians not to slow down the development of AI-based innovations prematurely by imposing too tight regulations. “If the government implements restrictions and bans too early, a lot of potential will remain untapped. Consequently, we need a flexible policy framework that minimizes risks but does not block opportunities.” He calls for application-related regulation. After all, whether AI makes medical diagnoses or optimizes technical processes in a factory are two very different things.
