Manh Ho

Recent advances in machine learning and neuroscience, the abundances of sensors and emotion data have endowed machines with ability to detect, classify, and interact with human emotions. This study philosophize about affective computing devices by examining emotional artificial intelligence (AI) through the lens of affective artifacts, focusing on as personal AI assistants, work-related emotion trackers, and emotional AI toys and robots. Drawing from philosophical accounts on affective artifacts and empirical evidence on human-AI interaction, we argue how emotional AI technologies fulfill the criteria of being affective artifacts from the four key properties: underlying algorithmic complexity, artifactual nature, semi-agentive capabilities, and semi-interpersonal interactions. Furthermore, recognizing that the underlying algorithmic complexity renders these technologies easily manipulated for emotion-related tasks, we also analyze how these tools help us regulate emotions and influence self-perception, interconnecting with the maintenance and construction of our sense of self and identity, blurring the lines between the infosphere and lived experience.

This study analyzes 324 online news articles published by five major Vietnamese state newspapers between 2016 and 2022. Besides rating the tone of each article as being positive, negative, or neutral, we also subject these articles to a rigorous thematic coding scheme; each article is read and coded nominally for whether it mentions AI-related topics such as AI ethics, risks, development, investment, AI public events, policies, etc. The analyses show the majority of news reports on AI are event-related, with 194 articles (59.8%) published about 151 unique AI-related events. Concurring with the literature on AI-related narratives, we conclude that industry actors are the main drivers of media narratives about AI. Hence, it is not surprising that the three major narratives are first, the prevailing optimism regarding AI development and application; second, the emphasis on AI development as one of the key contributors to the national stature; and third, the shallow discussion of AI ethics. We find the number of negative news items was dismissed at 4, equivalent to merely 1.23%, whereas that of neutral news items was 47, or 14.5%. Problematically, there is a stark shortage of editorial, op-ed, and in-depth feature articles explaining various ethical and social issues related to AI deployment. The AI-related ethical issues, such as privacy violations, gendered bias, etc., are merely listed as talking points and are being overshadowed by the narrative of ‘efficiency,’ ‘cybersecurity,’ ideological security,’ and ‘strategic advantage.’ Moreover, we also provide an explanation for the limited attention that AI-related ethical issues receive from government newspapers in Vietnam, attributing it to the lack of a philosophical foundation for the discussion on AI ethics and the optimistic view on AI of the Vietnamese state's journals is tied to the way the Vietnamese government is gaining legitimacy, that is, through economic developments.

Recent advances in machine learning and neuroscience, the abundances of sensors and emotion data have endowed machines with ability to detect, classify, and interact with human emotions. This study philosophize about affective computing devices by examining emotional artificial intelligence (AI) through the lens of affective artifacts, focusing on as personal AI assistants, work-related emotion trackers, and emotional AI toys and robots. Drawing from philosophical accounts on affective artifacts and empirical evidence on human-AI interaction, we argue how emotional AI technologies fulfill the criteria of being affective artifacts from the four key properties: underlying algorithmic complexity, artifactual nature, semi-agentive capabilities, and semi-interpersonal interactions. Furthermore, recognizing that the underlying algorithmic complexity renders these technologies easily manipulated for emotion-related tasks, we also analyze how these tools help us regulate emotions and influence self-perception, interconnecting with the maintenance and construction of our sense of self and identity, blurring the lines between the infosphere and lived experience.