Abstract
We here present some opinions about the challenge of understanding what the newer forms of artificial intelligence (AI) are, and how they are similar to or different from human intelligence (HI)in its many manifestations. Our central thesis is this: now that the inner workings of an AI may consist of a huge network, and billions of weight parameters, we might really benefit from a research framework that studies these new intelligences using the array of tools that have been developed to assess dimensions or aspects of intelligence in humans and in animals. We feel that in the current model computer scientists seek to adapt the ideas and nomenclature of psychology, or psychologists try to run black box AI systems through various kinds of " maze tasks. " While these efforts are aimed in the right direction, they are unlikely to birth a new science, which, we argue, is now needed. In this note we address several aspects of the problem, and propose that ways be found for psychologists and computer scientists to work together, respecting the strengths and weaknesses of their own disciplines, to develop much needed measuring instruments for AI, and its relation to human intelligence. We also offer some preliminary ideas about which dimensions of intelligence may prove salient when extending our notions of human or animal intelligence, to machines. This paper presents no new experimental or theoretical results.