Buckner, Cameron
(2019)
The Comparative Psychology of Artificial Intelligences.
[Preprint]
Abstract
The last five years have seen a series of remarkable achievements in Artificial Intelligence (AI) research. For example, systems based on Deep Neural Networks (DNNs) can now classify natural images as well or better than humans, defeat human grandmasters in strategy games as complex as chess, Go, or Starcraft II, and navigate autonomous vehicles across thousands of miles of mixed terrain. I here examine three ways in which DNNs are alleged to fall short of human intelligence: that their training is too data-hungry, that they are vulnerable to adversarial examples, and that their processing is not interpretable. I argue that these criticisms are subject to comparative bias, which must be overcome for comparisons of DNNs and humans to be meaningful. I suggest that AI would benefit here by learning from more mature methodological debates in comparative psychology concerning how to conduct fair
comparisons between different kinds of intelligences.
Monthly Views for the past 3 years
Monthly Downloads for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[feed]](http://philsci-archive.pitt.edu/style/images/feed-icon-32x32.png){kind=icon}