On Trust in the Foreground: Keynote Talk at the 2nd Workshop on Trustworthy Self-Organizing Systems (TSOS 2011)

摘要

It is the case that a growing number of trust models exist, each seemingly tailored to specific tasks and needs. In a great many cases, the trust models are intended to help people make decisions in uncertain circumstances or with a limited amount of knowledge. In almost all cases, this 'help' is in the form of an already made decision based on various calculations that may or may not be opaque to the user who is intended to be helped. Of course, there's nothing inherently wrong with this, but in some situations, we'd like the user to be somewhat more involved in the decision making process. At the very least, they should understand why systems suggest that others are trustworthy or not. The work of Natasha Dwyer explores, for example, the idea that, when people are interacting through networks, what is required is for them to be allowed to make their own trust decisions through information given by the system, not by the system telling them who to trust. This is the idea of Trust in the Foreground. Nowhere is this more important than in the ever-increasing situation of personal devices such as smart phones and tablets. These devices are explicitly aimed at a relationship between them and the user. This relationship is usually seen as unidirectional -- the user trusts the device. We are exploring how it can be looked at in the opposite direction -- device considering user, advising on external situations and internal requirements, and allowing the user to make up their own mind given this advice. We call this aspect of Foreground Trust 'Device Comfort'. This talk will examine the idea of Trust in the Foreground and why it is so powerful and necessary, before delving into Device Comfort as an example, and extending it further to the idea of 'Thing Comfort' in Internet of Things and Critical Infrastructure Interdependency models.