More Human Than Human: Does The Uncanny Curve Really Matter?

Post on 27-Jan-2015

109 views 2 download

Tags:

description

Presentation from HRI2013 Workshop on Design of Humanlikeness in HRI from uncanny valley to minimal design (in conjunction with 8th ACM/IEEE International Conference on Human-Robot Interaction, HRI2013). Full paper can be accessed from: https://www.academia.edu/7419951/More_Human_Than_Human_Does_The_Uncanny_Curve_Really_Matter

Transcript of More Human Than Human: Does The Uncanny Curve Really Matter?

Does the Uncanny Curve Really Matter?

Jakub Złotowski, Diane Proudfoot& Christoph Bartneck

Anthropomorphism - origins

Anthropomorphism

Why do we anthropomorphize?

Research aims

Anthropomorphism in HRISome of the factors affecting anthropomorphism:

● Embodiment● Movement● Verbal communication● Emotions● Intelligence● User related (motivation, cultural background)● Mere interaction

Outcomes of anthropomorphism● Empathy towards robots● Reactions similar to human-human

interaction● Social facilitation effect● Acceptance of robots

Uncanny valley

Uncanny cliff

Complexity of design

Directions for future research● Fixing terminology● Finding the inflection point● Investigating the entities that lie between

the deepest point of the uncanny valley and the human level

TerminologyShinwankan:

● familiarity ● likeability ● affinity ● eeriness ● empathy

Optimal level of human-likeness

Human-like entities● Sudden rise of likeability in the uncanny

cliff has not been demonstrated● Only humans were used as comparison

points● No entities were demonstrated to be

similar enough to humans for their likeability to increase after passing the uncanny curve

Dehumanization

Issues with previous studies

Conclusions● Anthropomorphism is common● Its impact on HRI is not well understood● Uncanny valley -> curve● Terra incognita of the graph● Where to place on the graph technology

enhanced humans?