In April I HarvardXR 2024 Augmented IntelligenceThere, I was astounded by the breadth of immersive technologies, diverse applications of AI, transformative theory, societal challenges, and emerging issues in design justice and education.
At this stage in my career, I often feel like an “AI immigrant.” Like many of my peers, including designers and academics, I am “overwhelmed” by the hype surrounding Artificial Intelligence (AI). However, for the younger generation, AI is seamlessly integrated into their education, careers, and everyday life, marking a major paradigm shift in most fields.
I am impressed and challenged by the diversity of AI tools and the evolution of the design process. The design environment and creative industries are undergoing profound change in every aspect: from frameworks, tools, and skills to processes, policies, justice, culture, and even power relations.
The integration of generative AI tools got me thinking about a few questions. How has the design process, typified by frameworks such as Double Diamond, evolved with the use of AI? Is there still room for human-led innovation within AI-driven design processes? And how can designers and researchers immerse themselves and invest in this AI transformation?
These questions are just the tip of the iceberg when it comes to AI, the creative process, and design in social contexts. But in this article, we’ll focus on three key insights:
- Curation is as important as creationCuration is as important as the creation itself.
- Distinguishing between design ideas and insightsDesign ideas do not directly reflect tastes or lifestyle aspects.
- Strengthening integration through imaginationUse your imagination to enhance the integration of new concepts.
Curation is as important as creation
The advent of AI and its integration into various disciplines has significantly changed the design process. A well-known model illustrating this transformation is the Double Diamond, introduced by the Design Council in 2003.
The model outlines an iterative process of convergence and divergence through four key stages—discover, define, develop, and deliver—that collectively represent a systematic approach to fostering creativity and innovation in design (Figure 2).
How has the design process for something like Double Diamond evolved with the integration of generative AI tools?
The integration of generative AI tools has significantly transformed the design process, especially during the early ideation stage. These tools enable individuals to generate design solutions and ideas more effectively and efficiently.
Twenty years later, one refinement of the traditional design process model is the Stingray model, which can be seen as an evolution of Double Diamond. The model has three stages: Train, Develop, and Iterate (Figure 3). Its purpose is to highlight how AI-powered tools can enhance the general creative process.
While designers and researchers still need to develop hypotheses, define the problem, and scope the project, the development and iteration phase is where we see the biggest change: with the help of AI-powered tools, designers can rapidly generate numerous concepts and develop numerous ideas.
This is followed by an iterative phase where ideas are validated using a combination of AI-driven and human-driven synthetic testing. This approach aims to incrementally refine the outcome to meet criteria for desirability, feasibility, and actionability.
Distinguishing between design ideas and insights
Design ideas and insights represent different stages within the design process. In the early stages, ideas often emerge from brainstorming and collaboration. These ideas can range from intuitive, diverse and practical, to novel and conceptual.
Insights typically emerge from a more structured, integrated process that leverages both qualitative and quantitative research. The goal of insights is to understand decision-making, human behavior, and reflections, and provide valuable learnings to key project stakeholders.
Design solutions, on the other hand, are the outcome of a design process that includes the stages of inspiration, ideation, prototyping, implementation, and refinement. These solutions integrate both ideas and insights.
However, there's a common misconception about generative AI tools. Many believe that the prompt-driven approach of these tools allows them to efficiently generate design solutions. In reality, these generative AI tools excel at generating a large number of early conceptual ideas without validation or experimentation.
Moving from an idea or insight to a solution is a time-consuming, research-heavy and validation-intensive process that often involves a lot of trial and error as different possibilities and scenarios are explored.
We have observed and used most generative AI tools, such as Hypersketch (Figure 4), DALL·E, and Midjourney, operate in a solution-driven mode. A key consideration is how they integrate observations from fieldwork and ethnographic research as key inputs before generating ideas. The prompts in these tools are based on inputs that reflect the needs and concepts of the designer, but often do not directly target the specific needs of the user.
Modern AI tools can take insights from research and generate a plethora of original ideas, but this abundance does not guarantee that designers or users have inherently good taste or the ability to help clients select and evaluate ideas according to their own preferences, desirability and lifestyle.
We often find ourselves lost in a sea of ideas, and we wonder if we can effectively sift through or curate these ideas to make meaningful progress or impact. This reflects the first point: curation is just as important as creation. Effective curation is shaped by life and work experiences, as well as collaborations.
It is very important to recognise the difference between a design idea and an insight, because generating a design idea does not necessarily translate directly to aspects of taste, lifestyle, branding or culture.
Strengthening integration through imagination
Many AI-powered tools and software have influenced the next level of competitive challenges, from creating individual products to building systemic platforms. These powerful AI-powered tools “communicate” with each other to produce impactful results. For example, beyond design tools, there are many academic research tools that have integrated AI-powered features. For example, ATLAS.ti, a computer-aided qualitative data analysis software, worked with Open-AI to apply AI-powered coding to identify people's language patterns, semantic analysis, and meaning making (Figure 5).
Just as consumers have personal preferences for electronic devices and operating systems like Apple or Android, each AI-powered product runs on its own platform. However, the integration of AI-powered tools has evolved to deliver optimal user-centric services and experiences.
Going forward, as designers and researchers face complex socio-technical challenges, the focus may shift from simply solving problems to emphasizing the formation of new perspectives that allow us to envision possible scenarios; how we can intentionally harness our imagination to futures and challenges to enhance the synthesis of new concepts with inclusive dimensions.
Summary: From artificial intelligence to predictive intelligence
In this article, we focus on AI, the creative process, and design in a social context, and we provide a preliminary look at three aspects: 1. curation is as important as creation; 2. distinguishing between design ideas and insights; and 3. enhancing integration through imagination.
The AI-powered exploratory Stingray model for innovation represents a transformation of the typical design process from Double Diamond. Is there still room for human-led innovation within AI-driven design processes, and how can designers and researchers immerse themselves and invest in this AI transformation?
While most academics and industry leaders interpret AI as “artificial intelligence,” we might broaden our understanding of the term to include “predictive intelligence.” This new interpretation encourages us to use AI to understand and solve challenges, adapt to change, and drive innovation.
To effectively leverage AI-powered tools, designers and researchers need to adopt the right mindset and skillset to think deeply and comprehensively about the nature of design, including considering the creative process, relevant frameworks, and design education. This raises important questions:
What future does humanity aspire to, and how can we achieve it with respect, meaning, justice and sustainability?
Overall, I hope to offer a nuanced view on the role of AI in design and advocate for a balanced approach where human creativity and AI capabilities are intertwined to address modern design challenges. Designers and researchers are encouraged to embrace AI not just as an efficiency tool, but as a partner in the creative process that helps push the boundaries of what is possible in design.
reference
