Humankind and machines – artificial intelligence in practice

Expectations of artificial intelligence (AI) are high. Driven by ever faster computers, via ever better algorithms and thanks to ever more available data, it is expected to provide crucial input to mastering social, scientific and economic challenges. At the same time, AI raises ethical and legal questions which can only be answered in a broad social discourse. How will coexistence, everyday life, work and the economy change? What is of benefit to humans beyond technical feasibility? What clichés and misunderstandings, or perhaps even fears, need to be dispelled?

Opportunities of AI-supported technology

When the boundaries between humans and machines become blurred, when machines are no longer merely tools but become quasi-colleagues making decisions to act, the rules of cooperation must be redefined. Science, business, politics and civil society all agree that AI has to work for and with humans, not against them. This is where DLR Projektträger (DLR-PT) comes in. For many years, we have been identifying the opportunities of AI-supported technology, developing innovative ideas for its design in order to make it socially acceptable and worthwhile for Germany and Europe. For example, we helped shape the Federal Government's "Artificial Intelligence Strategy" and are the project management agency responsible for AI topics for both the Federal Ministry of Education and Research (BMBF) and the Federal Ministry for Economic Affairs and Energy (BMWi).

Germany as an AI research location

Thanks to its strong basic research, Germany has a good starting position in AI, which the Federal Government is further intensifying by consolidating the existing competence centres for AI research, increasing the promotion of young scientists and broadly anchoring AI at universities. On behalf of the BMBF, DLR-PT is involved in the expansion of six national AI competence centres – including the German Research Centre for Artificial Intelligence (DFKI). One key question for us is how basic research results can be put to use as quickly as possible. Together with the BMBF, we have launched funding measures for the application of AI and machine learning in practice.

We are addressing fundamental questions about the understanding of AI systems. For example, within the framework of a guideline we developed, research is funded that makes artificial intelligence and machine learning more comprehensible. Because only when humans know how an AI makes decisions can they also evaluate them and, for example, determine who may be liable for an error. On behalf of the BMBF, we also bring AI experts from all over the world to Germany so that they can work together here on current research and development issues – for example, in the context of the "International Future Laboratories", which consist of up to 12 international researchers.

The ethics of machines

This raises the question of the need for ethics for machines. AI systems are designed to make decisions – from a multitude of options for action, they select the optimal solution according to their standards. This has consequences for humans. In the case of autonomous driving, for example, the question arises of making sure that an AI takes an ethically and morally justifiable decision in a critical traffic situation.

In the area of financial systems, too, AIs are increasingly being used to evaluate products and services. Thus algorithm-driven AI thus now has the potential to destabilise entire financial systems – what safeguards can be used here, and how can systems be controlled sustainably? These are questions to which we are developing answers. On behalf of the Baden-Württemberg Foundation, we are implementing the research programme "Responsible Artificial Intelligence", in which we accompany research projects that deal with the ethical use of AI for the good of the public. We have addressed further ethical, legal, social and political questions of AI, among others, in the context of the projects we initiated, ABIDA and its successor GOAL. ABIDA dealt with the question of which changes in the economy, services, government action or private life are triggered by the ever-increasing availability of data (Big Data). In the GOAL project network ("Governance by and through Algorithms"), computer scientists, legal scholars, ethicists, behavioural scientists and economists conduct interdisciplinary research on the role of technology as a means of effective, efficient and automated implementation and enforcement of social rules. Among other things, they deal with the question of what behavioural and regulatory effects algorithms have and how they can be used responsibly.

Digitalisation of the economy and Industry 4.0

As a building block of the German government's AI strategy, we have designed an innovation competition on behalf of the BMWi that aims to broadly promote the digital transformation of the economy. The areas of application in the 24 consortia we are currently supporting range from the construction industry and the health sector to quantum computing – including for small and medium-sized enterprises.

For the BMBF, we are implementing the funding measure "Research, Development and Use of Artificial Intelligence Methods in SMEs (KI4KMU)" and thus supporting SMEs in accelerating technology transfer from the pre-competitive area to practical application. In addition, on behalf of the BMWi, we have now helped set up 26 “Mittelstand-Digital Zentren” (SME Digital Centres) across Germany, where small and medium-sized enterprises (SMEs) are enabled to recognise the opportunities and challenges of the new technology, build up know-how and implement concrete applications with the help of over 50 AI trainers.