Helmut Leopold, Head of Center for Digital Safety & Security, AIT: "In order for AI to be widely used, fundamental problems still need to be solved." (Credit: AIT)
AI is a current hype, which will fundamentally change all areas of technology but especially our lives. Through AI we expect to solve problems with machines that previously seemed impossible. Whether it is image processing, text analysis or speech recognition, or the analysis of sensor and machine data, there are many expectations in artificial intelligence to become a miracle technology. In fact, there are even discussions that AI will be able to outstrip human intelligence in the future and that "super intelligence" can be expected.
However, if one takes a closer look at the topic, it quickly becomes clear that at first glance impressive results can be achieved for various problem solutions, but that very fundamental problems for a correctly functioning tool based on AI have not been solved yet. There are five basic challenges and related issues in international research that need to be addressed and resolved: modelability, verifiability, explainability, ethics and responsibility. Only if we have found answers to these questions, we will master the technology to the necessary extent, so that AI can serve us as an effective tool, which we can use responsibly. In this way we avoid being confronted with a technical unmanageable curse.
(Author: Christof Baumgartner; Computerwelt, 23.06.2020)