Interview and new material10 commandments of AI ethics - Is AI good or evil?

Questions for Prof. Dr. Petra Grimm "AI ethics: is AI good or evil?"

Prof. Dr. Grimm, is a world without artificial intelligence still conceivable today?

As long as there is electricity and no threat of a blackout, AI will not disappear from this planet. Whether in schools, the world of work, medicine or the military - AI systems are everywhere. Today, the question is no longer in which areas AI plays a role, but rather where it shouldnot play a role. It is important to bear in mind that AI is a huge consumer of resources and therefore not climate-neutral. Sustainable and responsible development of AI should therefore be a key objective for companies.

Does AI bring more opportunities or more risks for us humans?

That depends on how we design and use it today and in the future. AI can calculate probabilities and recognize patterns - depending on the context and design, it can do good or bad things. The quality of the training data, for example, is crucial: If it contains biases or stereotypes, the AI will also adopt them. Developers must therefore implement "ethics by design" from the outset - i.e. a value-oriented system design. Operators and users should also critically examine whether ethical standards are being adhered to when selecting AI tools.

The use of AI can have advantages and disadvantages for different areas of life or interest groups. What opportunities do you see for society through AI? And what risks do you see?

In the field of education, AI could ideally help to make learning more individual, collective and needs-oriented. However, the tools must respect students' privacy, protect their data and not use AI to assess performance. Furthermore, AI must not be humanized: It cannot learn, feel or empathize - it merely calculates probabilities. A "relationship" with a chatbot therefore always remains one-sided. We need to ask ourselves: in which contexts is AI useful - and where is it better not?

How does AI change our image of humanity when technology takes on more and more tasks that are actually "typically human"?

Human images are fundamental ideas about what constitutes a human being - often unconsciously, but suddenly very present when AI is used. Those with a deficient view of human nature ("machines are more neutral than biased humans") are more likely to use AI for decision-making, for example when scanning job applications. Those with a cooperative view of human beings see AI as a tool that checks formal criteria, while the final decision remains with the human being. It is a myth that AI is "human": it lacks corporeality, experience and self-awareness. There are currently two competing images of humanity: On the one hand, the optimization of performance and efficiency, and on the other, the emphasis on human values such as freedom and transparency.

Prof. Dr. Grimm, you helped develop the 10 commandments of AI ethics. Can you give us suitable examples of some of the commandments?

In view of the trend of adolescents increasingly using AI as a companion or advisor, the second commandment is key: "Be aware that AI cannot replace real friendships or love." Children share secrets with chatbots that they would not confide in friends or family. There is an urgent need for education and AI competence here - especially against the backdrop of increasing loneliness.

Equally important is the fourth commandment: "Think for yourself first before using AI and train your own skills." Generative AI such as ChatGPT, Grok or Mistral was not developed for educational purposes, but is penetrating schools. AI can help as a tutor - only when children have learned to think for themselves. Otherwise, we risk weakening problem-solving skills and critical thinking.

So how can we shape a successful life with AI?

Together, we should consider what a good and successful life with AI could look like. How should we as humans behave towards this development? We should neither mystify AI as a savior nor as a destroyer. It is crucial to shape it in a way that is compatible with democracy, to preserve our autonomy and to strengthen our critical powers of judgment. At the same time, the political and economic dominance of a few AI players must be democratically regulated. We cannot allow a handful of companies to determine the future of AI alone.

Prof. Dr. Grimm, thank you very much for taking the time to answer our questions!