Understanding machines
The computer scientist Katharina Zweig encourages discussion about artificial intelligence.
Who doesn't feel - at least from time to time - left behind when discussing topics such as artificial intelligence? With this book, computer science professor Katharina Zweig from the Technical University of Kaiserslautern is trying to change that. According to her, everyone can and should take part in current debates about digitization, because no in-depth understanding of computer science is required.
The author guides her readers through the "ABC of computer science" (algorithms, big data and computer intelligence) by explaining terms that we now encounter almost every day without ever having been properly defined, in a way that is understandable and based on many examples explained. Numerous illustrations adorn the chapters, which often clearly present complicated facts. The character "KAI" appears again and again, an artificial intelligence that accompanies the readers through the entire book.
Target Separation
Branch describes where algorithms are used - and what consequences this has. She gives a number of examples: Some airlines use algorithms to specifically address people who probably belong together so that they can choose another seat for a fee. In addition to moral concerns, the author names other tangible disadvantages that such an approach can have. Studies have shown that a family's emergency evacuation can be delayed if family members are not sitting together.
But the computer scientist emphasizes that the use of artificial intelligence can also have positive effects. Such systems could, for example, help doctors to identify tumors in radiological scans. The author does not believe that AIs will ever replace humans. In her view, the new technologies represent an additional tool that could make work easier in many areas.
The scientist keeps telling personal anecdotes, for example how she decided to switch to computer science as a biochemistry student. In another chapter, she reports that her husband was recently flashed on the Autobahn - and how a decision tree can be used to find out what fine he should expect.
Branche's experiences as a member of a commission of inquiry that has been debating the use of artificial intelligence for the German Bundestag since 2018 are particularly interesting. She also describes how she exchanged views with a scientist who, with his working group, developed an algorithm for the Austrian labor market service. The program is designed to assess the likelihood of a job seeker being placed. At first, Zweig was skeptical, but in the course of the discussion, the computer scientist convinced her that the algorithm could have positive effects.
Zweig has, among other things, developed a "risk matrix" that she uses to divide the use of algorithms into certain categories. She considers purchase recommendations to be harmless, but autonomous weapon systems to be questionable. The author proposes how to regulate the programs of individual classes. From their point of view, weapon systems shouldn't be automated at all.
Although the author offers her own views on many controversial issues, she also offers other perspectives. For example, whether there could ever be a strong AI (comparably intelligent as a human being) - if such a thing is even possible. All in all, Zweig encourages her readers to form their own opinions and stand up for them – even if they aren't experts.
