In his guest article “AI will not replace software engineers!” in the Frankfurter Allgemeine Zeitung (FAZ ), Alexander Pretschner, Chairman of the bidt Board of Directors and Professor of Software & Systems Engineering at the Technical University of Munich, explains which tasks cannot be performed by AI. Many of these tasks require decisions that require a deep understanding of specific needs and the differences between potential solutions. However, an AI could only “guess” these needs based on the training data. According to Pretschner, different human skills are still crucial for the development of a customised solution.
Ultimately, someone has to be able to express what exactly users want in a form that can be utilised by software. How is an AI supposed to be able to do this ex nihilo?
Prof. Dr. Alexander Pretschner To the profile
Alexander Pretschner emphasises that he sees great potential in the use of generative AI in software engineering, e.g. in the creation of standard code. However, a software engineer will still have to “check and assess whether the artefact created by generative AI is professionally and technically appropriate.”
to the Article (in German)