AUTHOR=Stoop Ruedi 

TITLE=Note on the Reliability of Biological vs. Artificial Neural Networks

JOURNAL=Frontiers in Physiology

VOLUME=Volume 12 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2021.637389

DOI=10.3389/fphys.2021.637389

ISSN=1664-042X

ABSTRACT=Various types of neural networks are presently widely used in diverse technical applications, not least because neural networks are known to be able to ’generalize’. The latter property raises expectations that they should be able to handle unexpected situations similarly successfully as humans do. Using fundamental examples, we show that in situations that they have not been trained for, artificial approaches tend to run into substantial problems, which highlights a deficit in comparison to human abilities. For this problematic - that seems to have obtained little attention so far - based on simple examples we provide a first analysis that exhibits some key features responsible for the difference between human and artificial intelligence.