Final Theses

Business Analytics, Digital Platforms, Hybrid Intelligence

Proactivity in Human-AI Systems

Situation

In the rapidly evolving field of human-AI interaction, proactivity has become an aspect of enhancing user experience and system efficiency. Proactive AI systems can anticipate user needs, provide timely assistance, and even initiate interactions, thereby creating more seamless and intuitive experiences. This is particularly relevant in domains like customer service, where conversational agents and voice assistants are increasingly integrated to support both customers and service agents.

 

For example, proactive voice agents can act as a third party in conversations, offering suggestions or solutions without explicit prompts. Such systems require high levels of autonomy and raise intriguing questions about agency, trust, and control. Understanding the balance between AI autonomy and human oversight is essential for designing systems that are both effective and acceptable to users. Additionally, determining when and how an AI should intervene involves complex reasoning strategies to ensure that assistance is appropriate and welcomed.

 

Exploring different types of proactivity and their implications not only advances theoretical understanding but also provides practical insights for developing AI systems that better meet user needs.

Objective of the Thesis

The overarching goal of this general topic direction is to examine the concept of proactivity in human-AI systems, with a particular focus on proactive conversational agents. Depending on your interests, the thesis can take one of the following directions:

 

                1.            Proactivity in Voice Agents as a Third Party in Conversations

                Focus: Investigate how voice agents can proactively participate in conversations between humans, such as support models for customer support agents in customer service.

                Objectives:

                •              Analyze existing proactive voice agent systems and their roles in multi-party conversations.

                •              Identify challenges and best practices in designing voice agents that effectively support human interactions without causing disruptions.

                •              Propose design guidelines or frameworks for implementing proactive voice agents in customer service settings.

                2.            Autonomy, Proactivity, and Agency Theory

                Focus: Explore the high levels of autonomy required for AI proactivity and how this relates to agency theory.

                Objectives:

                •              Examine the relationship between AI autonomy and user perceptions of agency, control, and trust.

                •              Analyze how agency theory can inform the design and deployment of proactive AI systems.

                •              Develop a conceptual model outlining the balance between AI autonomy and user oversight in proactive systems.

                3.            Reasoning Strategies for AI Intervention

                Focus: Study the reasoning strategies that proactive AI systems use to decide when and how to intervene in human activities.

                Objectives:

                •              Review existing reasoning mechanisms in proactive AI systems, including rule-based, machine learning, and hybrid approaches.

                •              Assess the effectiveness of different strategies in various contexts.

                •              Propose new or improved reasoning strategies that enhance the appropriateness and effectiveness of AI interventions.

                4.            Types of Proactivity in Conversational Agents

                Focus: Investigate the different types of proactivity exhibited by conversational agents and their impacts on user experience.

Objectives:

                •              Classify and define various forms of proactivity (e.g., informational, action-oriented, context-aware) in conversational agents.

                •              Analyze user responses and preferences related to each type of proactivity.

                •              Provide recommendations for selecting and implementing proactivity types based on application domains and user needs.

 

By conducting a structured literature review, you will synthesize existing research, identify gaps in knowledge, and contribute to the theoretical and practical understanding of proactivity in human-AI systems.

Your Profile

   •              Interest in Human-AI Interaction: A keen enthusiasm for exploring how humans interact with AI systems, particularly in conversational contexts.

                •              Research Skills: Ability to conduct comprehensive literature reviews and critically analyze academic papers.

                •              Analytical Thinking: Strong skills in conceptual thinking and the ability to develop theoretical frameworks.

                •              Independent Work Ethic: Self-motivated and capable of working independently while managing project timelines effectively.

                •              Communication Skills: Proficiency in scientific writing and the ability to articulate complex ideas clearly in English.

We offer

  •              Expert Supervision: Close guidance from researchers at our chair.

                •              Collaborative Environment: Opportunities to engage with a research team and possibly contribute to ongoing projects in related areas.

                •              Flexible Timeline: The ability to start at your convenience.

                •              Academic Contribution: The potential to publish your findings in academic conferences or journals, contributing to the field’s advancement.

Application

If you are excited about the prospect of exploring proactivity in human-AI systems and wish to contribute to this cutting-edge area of research, please submit:

 

                •              Your CV: Highlighting relevant academic achievements, research experiences, and skills.

                •              Academic Transcripts: Providing details of your coursework and grades.

                •              Motivation Letter: A brief statement (max. 200 words) expressing your interest in the topic and indicating which research direction(s) you are most interested in.

                •              Proposed Timeline: An outline of how you plan to approach the thesis and manage your time over the course of the project.

Level stage

Bachelor

Chair

Research Group Prof. Dr. Ivo Blohm

Persons

Marc Christopher Grau

To Detail
north