On the Trustee’s Side: Trust As Relational Capital - Trust Theory: A Socio-Cognitive and Computational Model - page 292

A general rule (that could be easily translated into an algorithm) regards the fact that the
trust capital of an agent (say Ag i ) increases when:
the number of other trusted agents (competitors) in the DN offering the solution to the given
task (or classes of tasks) decreases; and/or
the number of agents (delegators/clients) in the DN requiring the solution to the given task
(or classes of tasks) increases.
Following this analysis, the trustee should work to decrease the number of competitors (for
example, disconnecting the links in the network, reducing their reputation, and so on) and/or
he should work to increase the delegators (for example, connecting new ones, changing the
needs of the connected ones, and so on).
Let us consider what kind of strategies can be performed to enforce the other's dependence
beliefs and his beliefs about agent's competence .If Ag i is the potential trustee (the collector
of the trust capital) and Ag j is the potential trustor we can say:
i) Ag i can make Ag j dependent on him by making Ag j lack some resource or skill (or at least
inducing Ag j to believe so). He has to work on SPN T (Ag j ,g jk ) .
ii) Ag i can make Ag j dependent on him by activating or inducing in them a given goal (need,
desire) in which Ag j is not autonomous (Castelfranchi and Falcone, 2003) but is dependent
on Ag i (or in any case they believe so). In this case they have to find a way to include
in G j an additional g jk such that Ag j is dependent on Ag i for that goal (and they believe
that).
iii) Since dependence beliefs are strictly related to the possibility of others (for example Ag j )
being able to see the agent (for example Ag i ) in the network and to know their ability to
perform useful tasks, the goal of the agent who wants to improve their own relational cap-
ital will be to signal their presence and their skills ((Schelling, 1960), (Spece, 1973), (Bird
and Smith, 2005)). While to show his presence he might have to shift his position (either
physically or figuratively, for instance, by changing his field), to communicate his skills he
might have to hold and show something that can be used as a signal (such as an exhibition,
certificate, social status, proved experience, and so on). This implies, in the plan of actions
of the trustee, several necessary sub-goals to provide a signal. These sub-goals are costly
to achieve and the cost the agent has to pay to achieve them has to prove the signals to be
credible (of course without considering cheating by building signals). It is important to un-
derline that using these signals often implies the participation of a third party in the process
of building trust as a capital: a third party which must be trusted (Falcone and Castelfranchi,
2001). We would say the more the third part is trusted in society, the more expensive will
it be for the agent to acquire signals to show, and the more successful these signals will
be at increasing the agent's relational capital. Later we will see how this is related to the
process of transferring trust from one agent to another (building reputation). Obviously
Ag i 's previous performances are also 'signals' of trustworthiness. And this information is
also provided by the circulating reputation of Ag i ((Conte and Paolucci, 2002), (Jøsang and
Ismail, 2002)).