Improving robot transparency: real-time visualisation of robot AI substantially improves understanding in naive observers

Research output: Contribution to conferencePaper

  • 1 Citations

Abstract

Deciphering the behaviour of intelligent others is a fundamental characteristic of our own intelligence. As we interact with complex intelligent artefacts, humans inevitably construct mental models to understand and predict their behaviour. If these models are incorrect or inadequate, we run the risk of self deception or even harm. Here we demonstrate that providing even a simple, abstracted real-time visualisation of a robot’s AI can radically improve the transparency of machine cognition. Findings from both an online experiment using a video recording of a robot, and from direct observation of a robot show substantial improvements in observers’ understanding of the robot’s behaviour. Unexpectedly, this improved understanding was correlated in one condition with
an increased perception that the robot was ‘thinking’, but in no conditions was the robot’s assessed intelligence impacted. In addition to our results, we describe our approach, tools used, implications, and potential future research directions.

Conference

ConferenceIEEE RO-MAN 2017
CountryPortugal
CityLisbon
Period28/08/171/09/17
Internet address

Fingerprint

Transparency
Visualization
Robots
Video recording
Experiments

Cite this

@conference{ba5404c352c24af88529e07ad9bbcbcd,
title = "Improving robot transparency: real-time visualisation of robot AI substantially improves understanding in naive observers",
abstract = "Deciphering the behaviour of intelligent others is a fundamental characteristic of our own intelligence. As we interact with complex intelligent artefacts, humans inevitably construct mental models to understand and predict their behaviour. If these models are incorrect or inadequate, we run the risk of self deception or even harm. Here we demonstrate that providing even a simple, abstracted real-time visualisation of a robot’s AI can radically improve the transparency of machine cognition. Findings from both an online experiment using a video recording of a robot, and from direct observation of a robot show substantial improvements in observers’ understanding of the robot’s behaviour. Unexpectedly, this improved understanding was correlated in one condition withan increased perception that the robot was ‘thinking’, but in no conditions was the robot’s assessed intelligence impacted. In addition to our results, we describe our approach, tools used, implications, and potential future research directions.",
author = "Wortham, {Robert H} and Andreas Theodorou and Bryson, {Joanna J}",
year = "2017",
month = "8",
day = "31",
language = "English",
note = "IEEE RO-MAN 2017 : 26th IEEE International Symposium on Robot and Human Interactive Communication ; Conference date: 28-08-2017 Through 01-09-2017",
url = "http://www.ro-man2017.org/site/",

}

TY - CONF

T1 - Improving robot transparency

T2 - real-time visualisation of robot AI substantially improves understanding in naive observers

AU - Wortham, Robert H

AU - Theodorou, Andreas

AU - Bryson, Joanna J

PY - 2017/8/31

Y1 - 2017/8/31

N2 - Deciphering the behaviour of intelligent others is a fundamental characteristic of our own intelligence. As we interact with complex intelligent artefacts, humans inevitably construct mental models to understand and predict their behaviour. If these models are incorrect or inadequate, we run the risk of self deception or even harm. Here we demonstrate that providing even a simple, abstracted real-time visualisation of a robot’s AI can radically improve the transparency of machine cognition. Findings from both an online experiment using a video recording of a robot, and from direct observation of a robot show substantial improvements in observers’ understanding of the robot’s behaviour. Unexpectedly, this improved understanding was correlated in one condition withan increased perception that the robot was ‘thinking’, but in no conditions was the robot’s assessed intelligence impacted. In addition to our results, we describe our approach, tools used, implications, and potential future research directions.

AB - Deciphering the behaviour of intelligent others is a fundamental characteristic of our own intelligence. As we interact with complex intelligent artefacts, humans inevitably construct mental models to understand and predict their behaviour. If these models are incorrect or inadequate, we run the risk of self deception or even harm. Here we demonstrate that providing even a simple, abstracted real-time visualisation of a robot’s AI can radically improve the transparency of machine cognition. Findings from both an online experiment using a video recording of a robot, and from direct observation of a robot show substantial improvements in observers’ understanding of the robot’s behaviour. Unexpectedly, this improved understanding was correlated in one condition withan increased perception that the robot was ‘thinking’, but in no conditions was the robot’s assessed intelligence impacted. In addition to our results, we describe our approach, tools used, implications, and potential future research directions.

M3 - Paper

ER -