Related Experiment Video
Updated: Jul 9, 2025

The Use of the Puzzle Box as a Means of Assessing the Efficacy of Environmental Enrichment
Published on: December 29, 2014
You can't play 20 questions with nature and win redux
Bradley C Love1, Robert M Mok2
1Experimental Psychology, University College London, London, UK b.love@ucl.ac.uk https://bradlove.org/.
Abstract:
An incomplete science begets imperfect models. Nevertheless, the target article advocates for jettisoning deep-learning models with some competency in object recognition for toy models evaluated against a checklist of laboratory findings; an approach which evokes Alan Newell's 20 questions critique. We believe their approach risks incoherency and neglects the most basic test; can the model perform its intended task.
Related Concept Videos
Limits to Natural Selection
Competition
What is Natural Selection?
Nature and Nurture
Natural Selection and Adaptation
Beyond physical adaptations,...
Robbers Cave

