AAAI Publications, 2018 AAAI Spring Symposium Series

Font Size: 
From Algorithms to Heuristics: Will Androids Ever Make Freudian Slips?
Sadeq Rahimi

Last modified: 2018-03-15

Abstract


This paper presents a conceptual framework for understanding the importance and relevance of human cognitive biases in development of more effective and seamless human-AI interactivity. A categorization model is proposed for distinguishing two general types of human biases: 1) biases that are primarily determined by neurobiological hardwiring, and 2) those that are determined and transmitted across generations by cultural and symbolic structures of thought and affect. It will then examine the meaning and implications of the suggested types of biases, specifically in relation to human-AI interaction.

Keywords


Cognitive bias; culture; AI;

Full Text: PDF