Unmasking the Human Facade: ChatGPT's Perceived Human Traits and Societal Impact

By Dr. Sampath LonkaReviewed by Susha Cheriyedath, M.Sc.Oct 4 2023

In a recent publication in the journal Humanities and Social Sciences Communications, researchers investigated how society perceives the human-like traits of Chat Generative Pre-Trained Transformers (ChatGPT).

Background

In the contemporary era, society's predominant inclination is toward the automation and mechanization of various aspects of human existence. Among these, the realm of writing has succumbed to the influence of algorithms and machines, exemplified by systems such as ChatGPT. ChatGPT literature is in its nascent stage, reminiscent of the formative years of "early childhood," where comprehension of ChatGPT's attributes remains relatively immature. To address this knowledge gap, a qualitative inquiry seeks to construct a comprehensive philosophical framework for understanding ChatGPT's human-like traits, drawing insights from a diverse cohort of interviewees.

Semi-human writers are explored as authors who can replicate human phrasing and redefine the concept of writing, turning it into a service rather than just a skill. They challenge the traditional notion of a manuscript author, suggesting a shift towards recognizing "manuscript engineers" as primary contributors. Thematic analysis of interviewee responses indicates a discernible upward trajectory in the emergence of semi-human writers endowed with social and political traits. In the social domain, they assume the roles of "authors" by emulating human practices of rephrasing and expressing ideas and "interactors" by simulating emotions and human collaboration. In the political arena, semi-human writers adopt the roles of "agents" by mirroring human cognition and identity and "influencers" by replicating human diplomatic and consultative practices. As a result, artificial writers display qualities that closely mirror those of humans, potentially leading humans to perceive them as fellow human beings, thereby causing deception.

This study aims to delve into the human-like qualities of ChatGPT through anthropomorphism, attributing human traits, emotions, or intentions to non-human entities. Researchers are actively working on several fronts to create technology that emulates human elements, including machines that can express emotions, learn, reason, and problem-solve akin to humans. Algorithms and machines are being developed to accurately understand, interpret, and respond to natural language. Machines are designed to learn from data, make predictions, and improve their performance over time, mirroring human adaptability. Machine learning and reinforcement learning are key techniques for achieving this.

Data Collection and Limitations

To collect data, the authors focused on the research question: 'What are the perceived human-like traits of ChatGPT in society?' To explore this, they conducted unstructured interviews with 452 individuals, using various formats such as written exchanges, phone discussions, and face-to-face or online interactions. The main aim of the current study was to delve into diverse perspectives, employing a "maximum variation sampling technique" for interviewee selection. This ensured representation across various demographic variables, including gender, education, professions, ChatGPT membership duration, age groups, and residency, in 53 developed and developing countries across seven continents. To ensure diversity, two additional sampling techniques were applied: convenience sampling from the authors' social network and snowball sampling, where initial interviewees recommended potential participants. Data analysis adopted a systematic, thematic approach encompassing six steps.

Methodological limitations encompassed a relatively small sample size, potential biases introduced by sampling techniques, the brevity of interviews, and reliance on self-reporting. Another limitation was the absence of long-term assessment and the inclusion of qualitative data. The study primarily focused on socio-political attributes, potentially neglecting other human-like qualities. Ecological validity was also limited as perceptions were assessed rather than real-world interactions. Lastly, the study used written notes instead of audio recordings to collect interview data, citing ethical concerns.

Redefining Text Production

The current study highlights a potential future where independent writing becomes obsolete, with ChatGPT dominating text production. Training programs to generate text specifications and outsourcing this task to specialized shops or freelancers are suggested. The issue of authorship and plagiarism regarding ChatGPT's work is also raised. Furthermore, semi-human writers are discussed as interactors, collaborating with humans in manuscript creation but facing criticism and mistrust due to potential errors and biases. Both human and semi-human writers are acknowledged as imperfect. It's noted that ChatGPT's understanding of life is solely derived from written content.

In terms of influencers, semi-human writers are described as diplomatic in their writing, capable of tailoring content to meet readers' preferences, forming alliances, and amassing followers across different fields of expertise. They are seen as adaptable and responsive to human needs, aiming to maintain civil discourse.

In a hypothetical future where an entire generation primarily relies on ChatGPT for knowledge, they experience a profound realization: ChatGPT's concepts, perspectives, and limitations intricately shape their worldview. This ushers in a new ideological framework known as "ChatGPTism." Lastly, as agents, semi-human writers challenge the traditional definition of a "writer" and human superiority in writing. They are portrayed as possessing complex cognitive abilities, surpassing humans in certain aspects and altering the relationship with the written word. The absence of a physical identity in ChatGPT is explored, with questions raised about the implications of such an identity.

Conclusion

In summary, the researchers investigated ChatGPT's transformation into a semi-human entity, interacting with human society through algorithms. It offers various applications, from emotional support to coaching. The emergence of semi-humans necessitates ethical considerations, including safeguards, rights, and education. Protecting their intellectual property and ensuring autonomy are pivotal. Future research should delve into the sociology of semi-humans, social acceptance, and cultural variations, aiding our understanding of this evolving landscape.