Michael Lissack, Hugo Letiche, and Brenden Meagher
The advent of large language models (LLMs) is transforming how researchers in design and science gather insights, draft manuscripts, and generate knowledge. This essay examines the dual challenge that emerges as science confronts LLM-based tools through a systems thinking lens: upholding rigorous scientific integrity while simultaneously evolving epistemic standards to accommodate AI-assisted inquiry. A synthesis of research and position papers demonstrates how LLM-based tools can be influential collaborators in design-oriented research—augmenting human creativity and analysis—while requiring new approaches to trust, authorship, and credibility. The concept of the augmented hermeneutic circle illustrates an iterative human-AI partnership that creates emergent understanding through complex feedback loops. The discussion addresses risks, including hallucinated facts, plagiarism, and misuse of LLM-based tools, arguing that knowledge production and trust mechanisms must co-evolve as interdependent elements within a complex sociotechnical system. Instead of treating LLM-based tools as mere tools or threats, designers and researchers are urged to see them as mirrors reflecting systemic patterns, compasses guiding complex information landscapes, and collaborators navigating emergent design spaces. The essay concludes by considering how design science and scholarly norms can adapt in tandem with these technologies, ensuring that the future of research remains innovative and credible within increasingly complex knowledge ecosystems.
Keywords: large language models, design research, complex adaptive systems, augmented hermeneutic circle, trust, epistemic evolution, relationality