Are We Trusting AI Too Much? Examining Human-Robot Interactions in the Real World

Explore a thought-provoking keynote address from the 2020 ACM/IEEE International Conference on Human-Robot Interaction, delivered by Dr. Ayanna Howard of Georgia Institute of Technology. Delve into the critical question of whether we are placing too much trust in artificial intelligence, examining real-world human-robot interactions. Gain insights into studies revealing people's tendency to overtrust AI systems, particularly in high-risk scenarios like healthcare. Learn about the potential consequences of this overtrust, especially when human judgment should take precedence. Discover methods for identifying and mitigating overtrust, with a focus on vulnerable populations such as children with disabilities. Explore fascinating topics including trust, bias, compliance, cognitive biases, and emotional reactions to robot mistakes. Understand the implications of these findings for the future of human-robot interaction and the ethical considerations surrounding our increasing dependence on intelligent systems.

Introduction
What is trust
Bias
Children with Disabilities
Humans are Science
Compliance
Trust
Biasing
We do it on purpose
The relationship with kids
The relationship with surveys
The outcomes are different
Cognitive biases
Two studies
Whats interesting
The Shooter Bias Test
Humans are robots
Violations of trust
Four conditions
Impact on trust
Emotional reactions when robots are wrong
The last study
Robot navigation
Robot mistakes
Risk