Hybrid intelligence processes and narratives in robotic manufacturing

This project investigates how different AI narratives—focusing on automation versus hybrid intelligence—impact students' perceptions about AI in the workplace. Early findings indicate that the type of narrative significantly influences their future professional outlook. To obtain these findings, we have employed a combined approach consisting of discussions, workshops, and surveys.

In the narrative workshops, students were divided into groups to explore two contrasting views on AI: one highlighting AI as a potential threat to jobs through automation, and the other presenting AI as a supportive partner in human-AI collaborations aimed at augmenting human skills..

This was followed by interactive technological workshops where students participated in practical activities, creating chatbots and programming robots. This hands-on experience helped them understand the real-world implications of the different AI narratives.

Additionally, surveys conducted before and after the workshops measured the impact of these narratives on students' perceptions of AI. This data collection was crucial for identifying which narrative frameworks positively shaped students’ views on AI’s role in professional settings.


Study procedure overview


CHI custom chatbot

For the project workshop, we developed a chatbot using OpenAI's Assistants API, tailored specifically to allow students to practice prompt engineering for website chatbots. This tool was used by 41 participants, who engaged in a total of 279 conversations, generating 1,104 messages.

This practical, hands-on experience not only deepened the students' understanding of chatbot functionalities and prompt engineering but also occurred within a structured setting. The extensive interaction with the chatbot provided significant insights into user behaviors and affirmed the benefits of such technologies in educational settings.

The chatbot incorporates several advanced features, including:

  • User authentication
  • Document uploading capabilities
  • Various actions, API authentication
  • Python runtimes (streaming)
  • OpenAI APIs (like RAG and Audio-to-X)
  • Persistent data storage
  • The ability to update assistant memory within a chat session.

Learn more about our chatbots and no-code platform here


Team

Michail Beliatis

Associate Professor

mibel@btech.au.dk
Herning, Innovatorium DBD-212

Ida Bang Hansen

Research Assistant

ida@mgmt.au.dk
2627, H206

Patrycja Julia Rogozińska

Research Assistant 

pjr@mgmt.au.dk
2627, H203

Jacob Sherson

Benedicte Ravn

Ph.d. Student

benedicteravn@btech.au.dk
Herning, Innovatorium DBD-207