The development and application of artificial intelligence (AI) are highly relevant current issues that affect the economy, science, politics and society in equal measure. The Centre for European Economic Research (ZEW), Mannheim, made its own contribution to the debate surrounding the opportunities and risks associated with this new technology with an event held on Thursday, 12 July 2018, as part of the series “First-Hand Information on Economic Policy”. Claudia Nemat, executive board member at Deutsche Telekom responsible for technology and innovation, was invited to deliver a speech on “Artificial Intelligence: A Wake-up Call for Europe” and provided the 120 guests in attendance fascinating, practical insight into the topic from the perspective of the biggest telecommunications firm in Europe.

Claudia Nemat gave an insight into KI from the perspective of Deutsche Telekom.
Claudia Nemat from Deutsche Telekom during her talk on artificial intelligence at ZEW.

Claudia Nemat opened her speech with the bold statement that AI is set to fundamentally change the way we work and think. That being said, AI is still far from being the “holy grail” that can solve all our problems. According to Nemat, who holds a degree in physics, this is because, up to this point, the only type of AI being implemented in practice is so-called “weak AI”, which develops algorithms based on previously supplied data. Currently, AI applications are in fact already capable of making decisions more quickly than the human brain or carrying out production techniques more precisely than the human hand. However, these applications have not yet succeeded in transferring intelligence from one specific area to another. AI that has so far been in use still does not make its own decisions when faced with uncertainty, but is rather largely a reflection of its creator.

“The performance of algorithms, such as those used to recognise patterns, can only be as good as the data it has previously been fed,” Nemat pointed out. One issue that has provoked heated discussion in this context is the procurement of personal data. According to Nemat, we need to find a way to make anonymised collected data accessible to companies whilst being transparent with consumers with regards to what is happening with their data. “Currently, AI is not self-aware, but is still very powerful in ways both good and bad. It is now up to us to learn how to shape AI,” said Nemat. There are currently many potential applications for this technology, particularly in the medical field, where it is already being used for the early identification of cancer cells and the diagnosis of rare diseases.

Compared to other countries, however, Germany lags far behind when it comes to further development of AI. With a total budget of five billion euros earmarked for AI research, the EU trails far behind Asia and the US, where almost double and quadruple this amount respectively is being invested in research initiatives related to AI. With this in mind, Nemat warned that Europe and Germany risk not only being at a great competitive disadvantage, but also of becoming dependent on other countries. Being forced into such a subordinate position involves the risk of overly strict regulation and could therefore serve as an obstacle to future innovation. “In Germany, we have a fantastic culture of engineering, but this needs to be reinvented in the wake of AI,” said Nemat.

According to Nemat, both Germany and Europe need to set additional goals when it comes to the development and implementation of AI technology. “Networking” was the order of the day, with Nemat calling for Europe’s industries to work in closer collaboration with one another in the future. This could help to make both Germany and Europe more attractive as a business location and also to acquire more international expertise. “AI is not a single machine; rather, it consists of countless different projects both big and small. The implementation of AI therefore requires a critical mass of people – and in addition we need a talent pool with a strong network of knowledge,” explained Nemat. She concluded her speech with a call for a European global market leader in the creation of industrialised algorithms.

Nemat’s speech was followed by a panel discussion, in which ZEW President Professor Achim Wambach expressed doubt over whether a market leader from Europe would also be able to achieve the desired economies of scale. Claudia Nemat countered, firstly, that economies of scale would be achieved through data and, secondly, that there would also be “economies of skill”, that would serve as magnets for talent, attracting further talent in turn and thus increasing exponentially in size. Achim Wambach pointed out that, while Europe has no problem with conducting basic research, the application of new technologies often proves more difficult. He therefore called for greater investment in application-oriented research, so that potential new applications for AI can be identified early and made accessible.

Nemat agreed with Wambach’s recommendation that more investment in research and development needs to come from companies themselves, though she also believed governments have a role to play: “For companies to become global leaders in technology, they need start-up funding. Governments ought to be providing these funds.” Following on from Nemat’s comment, the ZEW President pointed out that small and medium-sized enterprises (SMEs) in particular are still not in a position to finance AI technologies due to high costs.

The questions from the audience were a testament to the keen interest with which developments in AI are being observed in many sectors of the economy. Questions included what future collaboration involving AI might look like, with Nemat pointing to open source platforms as a particularly advantageous model, for example, for exchanging trained algorithms. Audience members also posed questions related to the changes in the job market as a result of digital transformation. “The amount of work out there is definitely not going to decrease,” reassured Nemat. What will be most challenging for workers, however, is “skill transformation”, which aims to provide workers with further training to prepare them for the challenges of the future labour market.

What areas related to the future of AI does future research need to tackle? Can AI be legally protected? What might the business model of a European global market leader look like? In Claudia Nemat’s opinion, what might modern immigration laws taking recent developments into account look like? These are just some of the questions posed over the course of the evening that show that the topic of AI also needs to be approached via a debate over its many possibilities – how to identify its potential applications early on and implement them effectively if Germany and Europe want to compete internationally in this arena.