Personalization and Realism in Human-AI Interaction

The Evolution of Human-AI Interaction: Personalization and Realism

Recent advancements in the field of human-AI interaction have significantly focused on enhancing the personalization and realism of AI-driven systems. The integration of Large Language Models (LLMs) into embodied conversational agents has opened new avenues for naturalistic social encounters in virtual environments. These agents are now capable of adapting their personas to influence social-emotional processing and behavior, thereby creating more engaging and realistic interactions. Notably, the manipulation of linguistic features in LLMs to align with personality traits has demonstrated the models' potential to craft persuasive content, which could impact user mindsets and well-being. Additionally, the study of perceived obesity levels in humanoid robots has shed light on how physical attributes of AI entities can affect human trust, emphasizing the importance of design considerations in robotics.

In the realm of data delivery optimization, understanding user preferences for visual, tabular, or textual data representation has become crucial for enhancing user experience. This research underscores the significance of personal traits in shaping data preferences and highlights the potential of LLMs to replicate these preferences, paving the way for more personalized data tools. Furthermore, the multidimensional measurement of photorealistic avatar quality of experience has provided insights into the subjective usability factors of avatars, revealing a linear relationship between avatar affinity and realism, which could inform future design implications for telecommunication systems.

Noteworthy Developments

  • Persona-based LLM Agents: The study on persona manipulation in LLM-controlled agents reveals significant impacts on social-emotional processing and engagement.
  • Persuasive Linguistic Features: Research on LLMs' linguistic adaptation to personality traits highlights their persuasive capabilities and potential ethical considerations.
  • Avatar Quality of Experience: The multidimensional assessment of photorealistic avatars offers valuable insights into user perceptions and design implications for telecommunication systems.

Sources

The influence of persona and conversational task on social interactions with a LLM-controlled embodied conversational agent

The Dark Patterns of Personalized Persuasion in Large Language Models: Exposing Persuasive Linguistic Features for Big Five Personality Traits in LLMs Responses

To What Extent Does the Perceived Obesity Level of Humanoid Robots Affect People's Trust in Them?

Optimizing Data Delivery: Insights from User Preferences on Visuals, Tables, and Text

A multidimensional measurement of photorealistic avatar quality of experience

Built with on top of