Call for papers
Premise: The Problem of Infosocial Order
In open and complex information societies including diverse agents having
different tasks to be carried out within time constraints, the problem of
cooperation arises in different contexts:
- User-agent interaction, where personal assistants need to be trustworthy
and congruent with the expectations of human users (Conte & Castelfranchi, 2000).
- Competitive contexts, like agent-mediated e-commerce, where software agents
are used to act in the interest and on behalf of their self-interested users
(Rao, 1998; Crabtree, 1998). This generates potential social dilemmas between
users' and others' interests which call for social solutions. Among
other aspects, a socially unacceptable strategy is sometimes found to produce
self-defeating effects in the long run (Crabtree, 1998).
- Teamwork, which requires individual responsibility for the common task,
and therefore commitment (Jennings, 1995) and social control (Kaminka and Tambe, 2000).
Hence, the necessity to build agent architectures which achieve coordination and
cooperation, solve potential conflicts, play intermediary roles, reconcile local
and global utility, etc. In short, both "human adequacy" and "strict engineering
reasons" (Schillo et al. 2000) raise problems of infosocial order.
Solutions to these problems may consist of conventions and/or norms or formal laws.
In the former, social order results from converging behaviour based upon shared
expectations. In the latter, social order is achieved by the fact that the agents'
deliberative processes make reference to shared prescriptions, which
may possibly be backed systems of incentives and sanctions (as is the case for
enforceable laws).
Open Questions
However, solutions pose new problems while solving others. In particular,
conventions and norms may be partial, local and even antagonist. Hence,
software agents are expected to take the same decision faced by humans in natural
societies, i.e. to detect and solve conflicts among incompatible norms, laws and
conventions). In addition, institutional laws imply a costly and not always
efficient system of centralised control. On the other hand, conventions do not
ensure innovation and do not allow for control. Therefore, a
number of questions need to be answered:
- What are the specific problems of social order in e-societies?
- What are the effects of the solutions implemented so far?
- Are there differences between norms, conventions and laws? What are their
relative costs and benefits?
- What about "enforcement" and "control"? To what extent existing solutions are
successful, and at which cost?
- Can norms and laws evolve? How to reconcile conventions and innovation?
- What about institutional competence, i.e. the capacity to recognise, execute
norms, solve conflicts among them, etc.?
- What kind of agent architecture do these phenomena require? What can be
solved with learning, adaptive agents, and when are deliberative architectures
needed? What about integrated architectures, where deliberative and dynamic
components interact? What about implementing moral and social emotions or sentiments?
- What about social monitoring and control, in comparison with institutional
enforcement and control?
Objectives
The workshop is intended to reach a large audience from different scientific and
computational fields, such as AI & Law, deontic logics, computational social sciences,
multi-agent systems, and e-commerce. The unifying perspective is provided by our
emphasis on agent-based computer simulation, organisations and institutions design
and management, agent-mediated interaction (with a special attention to e-commerce
and virtual markets), multi-agent systems, info-societies design, AI-based
collaborative design. The workshop is also aimed at promoting useful discussion.
To this aim, the workshop will consist of a set of symposia on specific topics
that will be established by the organisers according to the results of the reviewing
process.
Authors are strongly recommended to submit original, although possibly
incomplete, work and, if their papers are accepted to participate in a collective
discussion about both their own and others' contributions. Both theoretical and
empirical works from a wide range of disciplines are encouraged, including
AI & Law, logical philosophy, philosophy of law, AI and computer science,
cognitive science,
evolutionary biology and psychology, philosophy of mind, cognitive psychology,
sociology and social psychology, economics, anthropology. Formal-computational
and simulation-based works are encouraged on a wide range of topics:
- Conceptual analysis of conventions, laws, & institutions
- Natural and e-societies
- Organisation design and management
- Collective vs. individual vs. shared obligation, commitment and responsibility
- Cooperation, commitment and teamwork
- Coordination
- Trust and reputation
- Fraud, exchange and deception
- Altruism and reciprocity
- Social and moral emotions
- Emerging vs. designed institutions
- Distributed vs. centralised control
- Enforcing mechanisms
- Agent architecture: deliberative vs. adaptive
Workshop Program Committee:
Francesco Billari, Max Planck Institute for Demographic Research, Rostock, Germany
Jos‚ Carmo,
Technical University of Lisbon, Portugal
Jos‚ Castro-Caldas, Higher Institute of Labour and Business Studies, Lisbon,
Portugal
Aspassia Daskalopulu,
King's College London, UK
Chris Dellarocas,
MIT Sloan School of Management, USA
Frank Dignum, Utrecht University,
The Netherlands
Nick Gotts,
Macaulay Land Research Institue, Aberdeen, UK
Andrew Jones,
University of Oslo, Norway
Jeremy Pitt, Imperial College London, UK
Carles Sierra,
Artificial Intelligence Research Insititute, Barcelona, Spain
Raimo Tuomela,
University of Helsinki, Finland
Back to the