Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Beyond Words: Infusing Conversational Agents with Human-like Typing Behaviors

Jijie Zhou, Yuhan Hu

TL;DR

The paper addresses how to augment ChatGPT-like agents with human-like typing cues to improve naturalness, engagement, and trust in conversational AI. It proposes a design space for hesitation and self-editing, implemented in a Flutter-based platform that renders typing dynamics and edits in real time. In a pilot within-subject study with three agent variants (Blue baseline, Green hesitation, Red hesitation plus self-editing), the Red agent achieved the strongest perceived naturalness, human-likeness, and competence, though effects were not statistically significant. The findings highlight the potential of fine-grained typing behaviors to enrich multimodal dialogue while underscoring the importance of content quality, memory, ethics, and broader applicability.

Abstract

Recently, large language models have facilitated the emergence of highly intelligent conversational AI capable of engaging in human-like dialogues. However, a notable distinction lies in the fact that these AI models predominantly generate responses rapidly, often producing extensive content without emulating the thoughtful process characteristic of human cognition and typing. This paper presents a design aimed at simulating human-like typing behaviors, including patterns such as hesitation and self-editing, as well as a preliminary user experiment to understand whether and to what extent the agent with human-like typing behaviors could potentially affect conversational engagement and its trustworthiness. We've constructed an interactive platform featuring user-adjustable parameters, allowing users to personalize the AI's communication style and thus cultivate a more enriching and immersive conversational experience. Our user experiment, involving interactions with three types of agents - a baseline agent, one simulating hesitation, and another integrating both hesitation and self-editing behaviors - reveals a preference for the agent that incorporates both behaviors, suggesting an improvement in perceived naturalness and trustworthiness. Through the insights from our design process and both quantitative and qualitative feedback from user experiments, this paper contributes to the multimodal interaction design and user experience for conversational AI, advocating for a more human-like, engaging, and trustworthy communication paradigm.

Beyond Words: Infusing Conversational Agents with Human-like Typing Behaviors

TL;DR

The paper addresses how to augment ChatGPT-like agents with human-like typing cues to improve naturalness, engagement, and trust in conversational AI. It proposes a design space for hesitation and self-editing, implemented in a Flutter-based platform that renders typing dynamics and edits in real time. In a pilot within-subject study with three agent variants (Blue baseline, Green hesitation, Red hesitation plus self-editing), the Red agent achieved the strongest perceived naturalness, human-likeness, and competence, though effects were not statistically significant. The findings highlight the potential of fine-grained typing behaviors to enrich multimodal dialogue while underscoring the importance of content quality, memory, ethics, and broader applicability.

Abstract

Recently, large language models have facilitated the emergence of highly intelligent conversational AI capable of engaging in human-like dialogues. However, a notable distinction lies in the fact that these AI models predominantly generate responses rapidly, often producing extensive content without emulating the thoughtful process characteristic of human cognition and typing. This paper presents a design aimed at simulating human-like typing behaviors, including patterns such as hesitation and self-editing, as well as a preliminary user experiment to understand whether and to what extent the agent with human-like typing behaviors could potentially affect conversational engagement and its trustworthiness. We've constructed an interactive platform featuring user-adjustable parameters, allowing users to personalize the AI's communication style and thus cultivate a more enriching and immersive conversational experience. Our user experiment, involving interactions with three types of agents - a baseline agent, one simulating hesitation, and another integrating both hesitation and self-editing behaviors - reveals a preference for the agent that incorporates both behaviors, suggesting an improvement in perceived naturalness and trustworthiness. Through the insights from our design process and both quantitative and qualitative feedback from user experiments, this paper contributes to the multimodal interaction design and user experience for conversational AI, advocating for a more human-like, engaging, and trustworthy communication paradigm.

Paper Structure

This paper contains 39 sections, 1 equation, 5 figures.

Table of Contents

  1. Introduction
  2. Related Work
  3. Typing/Character Perception
  4. Self-correction
  5. Temporal Delay
  6. ChatGPT/LLM-based and Conversational AI applications
  7. Design
  8. Design Space
  9. Platform Implementation
  10. Design Parameters
  11. Temporal Parameters
  12. Self-editing Parameters
  13. Experiment
  14. Research Questions and Hypothesis
  15. Methodology
  16. ...and 24 more sections

Figures (5)

  • Figure 1: User Interface Screenshot: Demonstrating the chatbot's typing behavior with self-editing capabilities (left), and adjusting parameters in ChatGPT's response (right). In our user experiment, the parameters are pre-defined for each agent.
  • Figure 2: Design space and parameters of human-like typing behaviors with examples.
  • Figure 3: Results of questionnaire responses.
  • Figure 4: Results of the count of word entry, sentence entry, and interaction duration time by the user and by the agent.
  • Figure 5: Linear Regression on the count of word entry by the user (Y axis) and by the agent (X axis).