Date of Award

8-4-2023

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Information Science and Technology

Advisor(s)

Kevin Crowston

Keywords

Artificial intelligence;Attitudes;Emotions;Level of engagement;Salience

Abstract

With today’s advanced technological improvements, the usage of various Artificial intelligence (AI) applications is still growing. This dissertation seeks to uncover how the future of AI is perceived by the social media users through the lens of technological frames and how these perceptions shape users’ emotions, attitudes, and the level of engagement stemming from salience of the frames. This shaping process is considered a type of social influence. The concept of technological frame is defined briefly as the interpretations, assumptions, expectations, and knowledge that people have about technology. Prior research suggests frames affect emotions and attitudes, and emotions and attitudes in turn affect peoples’ level of engagement. Even though many questionnaires and interviews have been conducted to understand the public’s attitudes towards AI and their relevant beliefs and views, many of these endeavors were not grounded on a theory and overlooked the connections between frames and emotions towards AI. On the grounds of framing theory and affective intelligence theory, this work investigates technological frames expressed in social media conversations where many users freely share their most recent ideas. The specific focus is Reddit, a huge social media platform that attracts millions of users with diverse mentalities shaped by different backgrounds, prior beliefs, personal experiences and personalities, from various geographical locations, thereby bringing different segments of the public together. More specifically, a corpus consisting of 998 unique future of AI-related post titles and their corresponding 16,611 comments created between 2012 and 2022 by 671 unique Redditors (Reddit users) was analyzed by using computer-aided textual analysis comprising a BERTopic model, and two BERT text classification models, one for emotion and the other for sentiment analysis, supported by human judgment. Finally, relationships among technological frames, emotions, attitudes, and the number of comments were examined to test a research model. The findings showed different interpretations about the power of AI and concerns such as possible justice and ethical problems stemming from AI usage (e.g., lack of laws, and privacy, bias, discrimination issues). However, the general attitude towards the future of AI was slightly positive and the most common feeling was curiosity. Moreover, the findings confirmed the research model we proposed and showed that technological frames affect social influence. More specifically, for example, we found that Benefits frame is positively related to curiosity and positive attitude. This original work makes several main contributions. As a practical contribution, the findings of this analysis can enrich current public voice-centric explorations of perceived future impacts of AI such as its benefits and risks. Also, the exploration of drivers of social influence in the context of technology may be useful for building awareness in society to accelerate the deployment and development of technologies for good of society. This study also provides policy implications. Academia, industry and government communities can collaborate to support policy arrangements in the areas where misconceptions about the future of AI are widespread. As theoretical and methodological implications, we propose a theoretical model and harness computer-aided textual analysis, which is applicable to further research. Lastly, this work helps us understand human communication in a technology setting focusing on frames, emotions and attitudes which are several elements that make us social creatures and develops computational language technologies that can discern these social elements in social media text data, thus constituting a bridge that connects fields of information systems, computational science and empirical social research.

Download

Available for download on Sunday, August 04, 2024

Share

COinS