The Segmented Double Diamond. A theoretical framework for innovation in the era of AI. : Intersection between Artificial Intelligence and innovation management through the lens of uncertainty toward a context aware AI integration.

Sammanfattning

The integration of Artificial Intelligence (AI) into human-centered design processes presents both opportunities and challenges. Specifically, within established frameworks like the Double Diamond (DD) design process. This study addresses a critical gap in effectively integrating AI into the DD framework by focusing on AI's contextual awareness of the design process without reducing the cognitive load required from the human factors to communicate their context for AI. The primary aim is to expand the DD framework to facilitate this integration in a way that enhance AI-Human synergy, focusing on the fluctuating levels of uncertainty inherent in the design process. By employing a Design-Based Research (DBR) paradigm, the study combines deductive and abductive reasoning to explore how uncertainty influences each phase of the DD model and develop a theoretical framework and design a practical prototype rooted in the theoretical findings and interpretation in the existing literature. A literature review and integrative analysis were conducted to ground the research in existing theoretical frameworks focusing to identify non-technological challenges in AI integration. challenges that are related to users’ behaviors and experiences. Qualitative data were collected through expert interviews and workshops with design professionals, providing users insights about the prototype into the collaborative dynamics between humans and AI under varying uncertainty levels. The findings mainly around identifying uncertainty as a core category across the DD as a system and all its constituting elements including AI. it is found that uncertainty levels increase during divergent phases (Discover and Develop) and decrease during convergent phases (Define and Deliver) of the DD process, forced by the distinct type of activities designers perform during exploration and selection. Based on this observation, the study proposes a segmentation of the DD framework into twelve distinct sub-phases categorized by low, moderate, and high uncertainty. This nuanced segmentation enables a more alignment with many AI techniques that rely on uncertainty differences within a data set. Such information can be intuitively given by the human factor in real time documentation, which underscore the importance of human factor to create synergy with the technology. A conceptual solution of data collection is suggested aiming to facilitate real-time data communication between human and AI, using the segmented DD framework. The theoretical contribution of this research lies in the segmentation of the DD framework based on uncertainty fluctuations, offering a deeper understanding of the design process and informing more effective AI integration strategies. Practically, the proposed prototype demonstrates how meaningful data collection can enhance AI's contextual awareness, leading to improved collaboration and innovation outcomes in design teams. The study concludes that addressing uncertainty and context in AI-human collaboration is essential for harnessing the full potential of AI in creative design processes. The study acknowledges the need for more expertise to execute a fully functional prototype and test it. The study also suggests the need for further research to properly test and refine a fully functional prototype based in the segmented DD framework, to test its practicality and functionality in real time communication with AI.