Aarhus University Seal / Aarhus Universitets segl

Do robots mean what they say? Curious robots, intentional agency and the logics of experimentation

Publikation: KonferencebidragKonferenceabstrakt til konferenceForskning

Standard

Do robots mean what they say? Curious robots, intentional agency and the logics of experimentation. / Vejlin, Frederik.

2021. Abstract fra Mega Seminar, Sønderborg, Danmark.

Publikation: KonferencebidragKonferenceabstrakt til konferenceForskning

Harvard

APA

CBE

MLA

Vejlin, Frederik Do robots mean what they say? Curious robots, intentional agency and the logics of experimentation. Mega Seminar, 23 aug. 2021, Sønderborg, Danmark, Konferenceabstrakt til konference, 2021.

Vancouver

Author

Bibtex

@conference{b5a16319fb8b4958b24b7c538f943af5,
title = "Do robots mean what they say?: Curious robots, intentional agency and the logics of experimentation",
abstract = "In this talk, I explore two quite different attempts at implementing some form of intentional agency in social robots that I witnessed during my fieldwork in 2017 at the Hiroshi Ishiguro Laboratories (HIL) in Japan. In the first case, the researchers at HIL staged a demonstration of the android ERICA{\textquoteright}s Intention-Desire System (the I-D System) for a film crew from Al Jazeera, who visited the lab to record a documentary on social robots in Japan. In the second case, I describe a human-robot interaction experiment designed to evaluate whether the algorithmic implementation of intrinsically motivated curiosity in the humanoid Robovie would produce more interesting and humanlike interactions when compared to a learning system that attempts to replicate socially appropriate behaviour. I combine Alfred Gell{\textquoteright}s discussions of intentionality and the art nexus in Art and Agency (1998) with Andy Pickering{\textquoteright}s redescription of scientific experiments as dances of agency (e.g. 1995, 2012) to discuss the different configurations of intentionality and agency that emerged in the two situations. I argue that although both ERICA and Robovie might arguably exhibit some interactive forms of intentional agency, the combination of their respective algorithmic systems and the relational matrix in which these were deployed shape their agential possibilities in qualitatively different ways. I show how ERICA{\textquoteright}s behaviour proceeds from a logic of replication that inhibits full participation in the dance of agency, thus resembling Gell{\textquoteright}s figure of the {\textquoteleft}secondary agent{\textquoteright}, whereas Robovie{\textquoteright}s algorithmic curiosity enacted a logic of experimentation that enabled the robot to take the lead, improvise new {\textquoteleft}choreographies,{\textquoteright} and thereby perform as a {\textquoteleft}primary agent{\textquoteright}. Finally, I use this discussion to cautiously approach the question of whether social robots mean what they say and why this might not be the best question to begin with. ",
author = "Frederik Vejlin",
year = "2021",
month = aug,
day = "24",
language = "English",
note = "null ; Conference date: 23-08-2021 Through 25-08-2021",

}

RIS

TY - ABST

T1 - Do robots mean what they say?

AU - Vejlin, Frederik

PY - 2021/8/24

Y1 - 2021/8/24

N2 - In this talk, I explore two quite different attempts at implementing some form of intentional agency in social robots that I witnessed during my fieldwork in 2017 at the Hiroshi Ishiguro Laboratories (HIL) in Japan. In the first case, the researchers at HIL staged a demonstration of the android ERICA’s Intention-Desire System (the I-D System) for a film crew from Al Jazeera, who visited the lab to record a documentary on social robots in Japan. In the second case, I describe a human-robot interaction experiment designed to evaluate whether the algorithmic implementation of intrinsically motivated curiosity in the humanoid Robovie would produce more interesting and humanlike interactions when compared to a learning system that attempts to replicate socially appropriate behaviour. I combine Alfred Gell’s discussions of intentionality and the art nexus in Art and Agency (1998) with Andy Pickering’s redescription of scientific experiments as dances of agency (e.g. 1995, 2012) to discuss the different configurations of intentionality and agency that emerged in the two situations. I argue that although both ERICA and Robovie might arguably exhibit some interactive forms of intentional agency, the combination of their respective algorithmic systems and the relational matrix in which these were deployed shape their agential possibilities in qualitatively different ways. I show how ERICA’s behaviour proceeds from a logic of replication that inhibits full participation in the dance of agency, thus resembling Gell’s figure of the ‘secondary agent’, whereas Robovie’s algorithmic curiosity enacted a logic of experimentation that enabled the robot to take the lead, improvise new ‘choreographies,’ and thereby perform as a ‘primary agent’. Finally, I use this discussion to cautiously approach the question of whether social robots mean what they say and why this might not be the best question to begin with.

AB - In this talk, I explore two quite different attempts at implementing some form of intentional agency in social robots that I witnessed during my fieldwork in 2017 at the Hiroshi Ishiguro Laboratories (HIL) in Japan. In the first case, the researchers at HIL staged a demonstration of the android ERICA’s Intention-Desire System (the I-D System) for a film crew from Al Jazeera, who visited the lab to record a documentary on social robots in Japan. In the second case, I describe a human-robot interaction experiment designed to evaluate whether the algorithmic implementation of intrinsically motivated curiosity in the humanoid Robovie would produce more interesting and humanlike interactions when compared to a learning system that attempts to replicate socially appropriate behaviour. I combine Alfred Gell’s discussions of intentionality and the art nexus in Art and Agency (1998) with Andy Pickering’s redescription of scientific experiments as dances of agency (e.g. 1995, 2012) to discuss the different configurations of intentionality and agency that emerged in the two situations. I argue that although both ERICA and Robovie might arguably exhibit some interactive forms of intentional agency, the combination of their respective algorithmic systems and the relational matrix in which these were deployed shape their agential possibilities in qualitatively different ways. I show how ERICA’s behaviour proceeds from a logic of replication that inhibits full participation in the dance of agency, thus resembling Gell’s figure of the ‘secondary agent’, whereas Robovie’s algorithmic curiosity enacted a logic of experimentation that enabled the robot to take the lead, improvise new ‘choreographies,’ and thereby perform as a ‘primary agent’. Finally, I use this discussion to cautiously approach the question of whether social robots mean what they say and why this might not be the best question to begin with.

M3 - Conference abstract for conference

Y2 - 23 August 2021 through 25 August 2021

ER -