What is it about?

When it comes to collaborating with other individuals, we all have different preferences. And this does not change if instead of other humans, we are co-creating with an AI. In this paper, we propose a framework that aims to identify key dimensions of user-system interaction. We then suggest further developing this conceptual framework into a practical tool in the form of a mobile application through which users can customise their creative experience, adapt the co-creation environment to their preferences, and choose the features of the AI they wish to engage with in the creative process.

Featured Image

Why is it important?

With the increasingly relevant role played by AI in content creation, there is a pressing need to examine the terms of the balance between automation and agency in human-AI interactions. This is necessary to continue reaping the benefits of increased efficiency resulting from the automatization of repetitive and burdensome tasks, preserving at the same time the users’ sense of creative agency and control. The novel contribution of the framework proposed in the paper is to identify the key dimensions responsible for the modulation of agency distribution between users and AI in co-creative processes. Furthermore, the customization app that we propose to develop based on the framework has the potential not only to improve user engagement, but also to address the inherent variability of human preferences for collaboration, recognizing the importance of adapting the co-creation environment to the different needs of individual users.

Perspectives

In the creative community there is an increasing urgency to establish clear boundaries between the phases of the creative processes in which AI should or should not interfere. With this framework we aimed to break down the complex concept of “agency” into actionable dimensions and parameters to modulate the level of control that users want to have on the creative process. I hope that the paper will inspire developers to think more deeply about how to adapt the co-creative experience to different user needs and profiles.

Caterina Moruzzi
University of Edinburgh

Read the Original

This page is a summary of: A User-centered Framework for Human-AI Co-creativity, May 2024, ACM (Association for Computing Machinery),
DOI: 10.1145/3613905.3650929.
You can read the full text:

Read

Resources

Contributors

The following have contributed to this page