178 Environments of intelligence
claimed that human beings, viewed as behaving systems, are quite simple and
that the apparent complexity of their behaviour is a function of the complexity
of their environments (Simon 1969, 126f ), he was, perhaps unwittingly, stating
a counterfactual conditional: there is no proper way of viewing human beings in
isolation form their environments to begin with. In being intrinsically, irreducibly
and dynamically coupled with their environments, human beings also have a hand
in shaping the kind and complexity of these environments – which will, in turn,
play a constitutive role in making the mind a rich, complex and adaptive affair.
The human condition is a moving target.
Notes
1 In terms of methodology, the interviews were of qualitative, semi-structured kind that
is common to “expert interviews”, as typically employed in the exploratory phases
studies that seek to gather an overview of the knowledge available in some field that
is then inquired mostly with respect to its social aspects (see Bogner et al. 2009; Gäser
and Laudel 2009; Meuser and Nagel 2009). The relevant social aspects with respect
to computer science and related fields were scientists’ and engineers’ perceptions of
human-technology relations and, more specifically, their views of abilities and purposes of users. In line with the format of the expert interview, the interviews were not
structured by a standardised list of questions but by an outline that stated the interview’s informational aims in more general terms. Interviews lasted between one and
one and a half hours, and were conducted face to face, with the exception of one telephone interview. The interviews were not tape-recorded but protocolled by the interviewer, who in all cases was the author of this study.
2 I am primarily referring to research at the University of Sussex by Inman Harvey
et al. (1994) and Ezequiel Di Paolo (2003). My account also draws on the exposition
in Boden (2006, 1325–1327). For a general introduction to the field of evolutionary
robotics, see Nolfi and Floreano (2000). For more recent surveys of the field’s development, see Doncieux and Mouret (2014) and Nelson (2014).
3 The discussion of the cognitive robotics approach is based on interviews with the
chair of the Institute for Cognitive Systems (ICS) at Technical University of Munich,
Gordon Cheng, and two researchers at the same department, Karinne Ramirez-Amaro
and Stefan Ehrlich, all of whom shall be acknowledged here. The interviews were supplemented by a survey of literature and brief demonstrations.
4 This is why I believe it is wrong to dismissively characterise experiments in these
fields as “caricatures” of human features and behaviours on the grounds of the simplification and exaggeration implied (this criticism is brought forward by Suchman
2008). The aim of these experiments is not a mockery of the human condition, but an
experimental inquiry into the physical and behavioural features by which people recognise each other as human beings. Even granted that humanoid robots and embodied
conversational agents are caricatures, a good caricature always is a pointed (albeit not
naturalistic) representation of certain salient properties of its subject matter.
5 The two sequences mentioned are at 1:47 through 1:53 and 1:37 through 1:46 minutes
respectively in one part of the online video documentation of the Sociable Machines project, titled “Recognition of Affective Intent”, http://www.ai.mit.edu/projects/sociable/
movies/affective-intent-narrative.mov.
6 See, for example the blog debates in https://jira.secondlife.com/browse/SVC-4180?
and http://alphavilleherald.com/2009/05/is-your-second-life-woman-a-real-life-man.
html. As Derek Stanovsky (2004) argues, a metaphysical point can be made of personal identities in virtual reality being ‘fake’ by necessity – but only a few SL residents
are likely to self-consciously play upon that point.
claimed that human beings, viewed as behaving systems, are quite simple and
that the apparent complexity of their behaviour is a function of the complexity
of their environments (Simon 1969, 126f ), he was, perhaps unwittingly, stating
a counterfactual conditional: there is no proper way of viewing human beings in
isolation form their environments to begin with. In being intrinsically, irreducibly
and dynamically coupled with their environments, human beings also have a hand
in shaping the kind and complexity of these environments – which will, in turn,
play a constitutive role in making the mind a rich, complex and adaptive affair.
The human condition is a moving target.
Notes
1 In terms of methodology, the interviews were of qualitative, semi-structured kind that
is common to “expert interviews”, as typically employed in the exploratory phases
studies that seek to gather an overview of the knowledge available in some field that
is then inquired mostly with respect to its social aspects (see Bogner et al. 2009; Gäser
and Laudel 2009; Meuser and Nagel 2009). The relevant social aspects with respect
to computer science and related fields were scientists’ and engineers’ perceptions of
human-technology relations and, more specifically, their views of abilities and purposes of users. In line with the format of the expert interview, the interviews were not
structured by a standardised list of questions but by an outline that stated the interview’s informational aims in more general terms. Interviews lasted between one and
one and a half hours, and were conducted face to face, with the exception of one telephone interview. The interviews were not tape-recorded but protocolled by the interviewer, who in all cases was the author of this study.
2 I am primarily referring to research at the University of Sussex by Inman Harvey
et al. (1994) and Ezequiel Di Paolo (2003). My account also draws on the exposition
in Boden (2006, 1325–1327). For a general introduction to the field of evolutionary
robotics, see Nolfi and Floreano (2000). For more recent surveys of the field’s development, see Doncieux and Mouret (2014) and Nelson (2014).
3 The discussion of the cognitive robotics approach is based on interviews with the
chair of the Institute for Cognitive Systems (ICS) at Technical University of Munich,
Gordon Cheng, and two researchers at the same department, Karinne Ramirez-Amaro
and Stefan Ehrlich, all of whom shall be acknowledged here. The interviews were supplemented by a survey of literature and brief demonstrations.
4 This is why I believe it is wrong to dismissively characterise experiments in these
fields as “caricatures” of human features and behaviours on the grounds of the simplification and exaggeration implied (this criticism is brought forward by Suchman
2008). The aim of these experiments is not a mockery of the human condition, but an
experimental inquiry into the physical and behavioural features by which people recognise each other as human beings. Even granted that humanoid robots and embodied
conversational agents are caricatures, a good caricature always is a pointed (albeit not
naturalistic) representation of certain salient properties of its subject matter.
5 The two sequences mentioned are at 1:47 through 1:53 and 1:37 through 1:46 minutes
respectively in one part of the online video documentation of the Sociable Machines project, titled “Recognition of Affective Intent”, http://www.ai.mit.edu/projects/sociable/
movies/affective-intent-narrative.mov.
6 See, for example the blog debates in https://jira.secondlife.com/browse/SVC-4180?
and http://alphavilleherald.com/2009/05/is-your-second-life-woman-a-real-life-man.
html. As Derek Stanovsky (2004) argues, a metaphysical point can be made of personal identities in virtual reality being ‘fake’ by necessity – but only a few SL residents
are likely to self-consciously play upon that point.