Please use this identifier to cite or link to this item:
Title: A turing test derivative : understanding public trust in AI through machine learning
Authors: Xiong, Rui
Chaiprasit, Rairom
Keywords: Social sciences::Communication
Issue Date: 2022
Publisher: Nanyang Technological University
Source: Xiong, R. & Chaiprasit, R. (2022). A turing test derivative : understanding public trust in AI through machine learning. Final Year Project (FYP), Nanyang Technological University, Singapore.
Project: CS/21/043 
Abstract: Given the rapid advancement of AI technology and the importance of trust for technology adoption, this study aims to evaluate how much and in what aspects the public trusts AI. Since it is difficult to define and measure trust accurately, inspired by the Turing test, we replaced our original questions with less ambiguous ones: To what extent do people deem it possible and desirable for AI to replace human workers? From a database of over 1000 jobs, we extracted 100 jobs with minimally overlapping characteristics using K-medoids clustering and conducted an online survey about these jobs on a representative US sample. Results from our hierarchical regression show that the perceived possibility and desirability of replacing human workers with AI are low for jobs requiring sensory-motor abilities, physical abilities, cognitive skills, and management skills, but high for jobs requiring mechanical skills. This suggests that public trust in AI is limited to work that is mechanical in nature. Furthermore, our results indicate that while the public generally lacks trust in the competency of AI technology, it is perceived desirable to replace human workers with AI for jobs involving precise control of body parts and social interactions. Implications and limitations are discussed.
Schools: Wee Kim Wee School of Communication and Information 
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:WKWSCI Student Reports (FYP/IA/PA/PI/CA)

Files in This Item:
File Description SizeFormat 
CS21043_Understanding Public Trust in AI through Machine Learning_Fundamental Research.pdf
  Restricted Access
Quantitative Fundamental Research752.08 kBAdobe PDFView/Open

Page view(s)

Updated on Apr 16, 2024


Updated on Apr 16, 2024

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.