Year: 2025
Type: conference-paper
Abstract: For discussions in the scientific community, we question the Post-Selection protocols (also known as p-hacking) in works that received the Nobel Prize in Physics 2024 and the Nobel Prize in Chemistry 2024, as well as the current flood of Post-Selection in Artificial Intelligence (AI) and machine learning within many other disciplines of science and engineering. We hypothesize that the Post-Selection protocol has three mistakes: (a) missing a test, (b) hiding bad-looking data, and (c) exaggerating prediction accuracy. Nobel Prizes must not escape the well-known scientific laws: (1) Different initial guesses in nonlinear problems lead to different local minima. (2) Papers about nonlinear problems must explain how the authors avoided local minima. Evidence: None of the questioned papers discuss (1) and (2). Therefore, they are not scientifically ethical. However, the future of AI and machine learning seems to be bright.
Authors: Juyang Weng, Colin Schmidt, Dongshu Wang, Ming Xie
Language: English
Subfield: Safety Research
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W4415125780
FWCI:
Cites:
Cited by:
Related to: