This week’s time was spent mostly on literature reading and understanding the suggested theories. I notice that there are theories that are similar but I feel both can be true. For instance, Media Richness Theory (MRT) and Media Naturalness Theory (MNT), while MRT suggests that communication effectiveness is dependent the richness that the communication medium that promotes social cues allows, so text messaging allows less richness in comparison to voice telecommunication, in comparison to video conference, in comparison to face-to-face of which has the highest level of richness. Meanwhile MNT argues that it is not the matter of richness, but rather naturalness, how closely the communication environment mimics the natural in-person communication environment that the communication channel allows to create the degree of effectiveness. I find the two theories are essentially communicating the same phenomena, and the distinctions between the two are so small that it’s insignificant.

Another theoretical detail I was stuck on, one of them being how to interpret self-disclosure. Research has shown that people are more comfortable disclosing information to an AI due to the perceived anonymity. This allows individuals to interact with AI without the need for impression management, ultimately leading to the sharing of more intimate and vulnerable information. This phenomenon is particularly beneficial in healthcare settings, such as mental health, where individuals may feel more comfortable disclosing information that might otherwise be embarrassing when discussing it with a human due to fear of judgment, as explained by social desirability bias.

Our experiment focuses on the context of employment interviews, where self-disclosure is expected to be “performative,” with responses and behaviors being closely scrutinized. Both the interviewer and interviewee are aware of these performance expectations. Recognizing the expected differences in outcomes could help us develop a clearer narrative and theory, emphasizing that the context in which AI is utilized will affect the outcome. While AI interactions may yield one result in a particular context, they do not necessarily produce the same effect in another.