Review of the book “The Depths of Deception” – Spectrum of Science

The word vertigo means an imbalance or a lie, depending on the case, and in the case of a simulation, a simulator is not far from pretending to be something. With such associations consciously used, Dippel and Warnke systematically adjust the difference between experiment and simulation over the course of the book. They claim that there has been a real paradigm shift: with the advent of complex model calculations, the methodology of all physics has changed. While attempts were made in the past to describe nature using differential equations, more recently people have relied on computer simulations.

In doing so, Dippel and Warnke are doing a disservice to the real-world content assessment of the digital models. They believe, for example, that with climate models one is satisfied with the reasonableness of the predictions derived from them and pays attention only to the similarity of the results of different simulations. But you are actually testing a model by examining whether it correctly predicts current values ​​from previous measurements, and only then letting it loose in the future.

The central criterion for the importance of experience, reproducibility, is also relative in Cultural Anthropology. A comment from quantum physicist Marcus Arndt, which he made during a visit to the authors at the University of Vienna, serves as proof: “We are the only ones who can perform such experiments. Nobody can reproduce it. Here the authors confuse the justified pride of the experimental physicist A head start in experimental technology With what is supposed to always be exclusive to her results. What no one else can repeat continually does not last in physics.

The unflinching pun on simulation and deception leads the authors to claim that the Turing test—a thought experiment pitting a human and an AI against one another—rewards, quite literally, “a fraud, because when an AI successfully competes to… impersonate a human, it has won.” It should be called smart.” The test’s inventor, Alan Turing, responded to this objection at the time: You can only tell whether a person is intelligent by their behaviour.