AI-Assisted Bio-Inspired Design: Evaluating the Effectiveness of Directed Intuition in Large Language Model Outputs
Journal of Mechanical Design · 01.12.2026 · 148 (12)
DOI: 10.1115/1.4072315 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Mechanical Design, online erschienen am 22.07.2026. Verfasst von Siyi Xiao und 4 weiteren Autor:innen. Die Arbeit nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Bio-inspired design is a well-established approach for generating creative ideas, and recent advances in large language models (LLMs) offer new opportunities for exploring such designs. Many studies have developed complex frameworks to integrate LLMs with bio-inspired design. One of the simplest methods a designer may use when first exploring bio-inspired design is through the directed intuitive approach, which instructs the designer to consider how nature might solve a given problem. While prior research has studied the impact of the directed intuitive method on human solutions, its effect on LLM-generated solutions remains unclear. This study investigates whether the directed intuitive method can lead to more effective LLM-generated solutions and whether the solutions are comparable to human-created solutions. A 2 × 2 × 2 factorial experiment was conducted using three factors: participant type, instruction type, and problem type. Results show that directing the LLM to “consider nature” increases the novelty and variety of the generated solutions, although the LLM still produces fewer novel ideas than humans. No significant differences were observed between human and LLM solutions in terms of variety and quality.
Zitieren
APA
Xiao, S., Hassan, H. I., Layton, A., Linsey, J., & McAdams, D. A. (2026). AI-Assisted Bio-Inspired Design: Evaluating the Effectiveness of Directed Intuition in Large Language Model Outputs. Journal of Mechanical Design, 148(12). https://doi.org/10.1115/1.4072315
Vancouver
Xiao S, Hassan HI, Layton A, Linsey J, McAdams DA. AI-Assisted Bio-Inspired Design: Evaluating the Effectiveness of Directed Intuition in Large Language Model Outputs. Journal of Mechanical Design. 2026;148(12). doi:10.1115/1.4072315
BibTeX
@article{Xiao2026,
author = {Xiao, Siyi and Hassan, Hadear Ibrahim and Layton, Astrid and Linsey, Julie and McAdams, Daniel A.},
title = {AI-Assisted Bio-Inspired Design: Evaluating the Effectiveness of Directed Intuition in Large Language Model Outputs},
journal = {Journal of Mechanical Design},
year = {2026},
volume = {148},
number = {12},
doi = {10.1115/1.4072315}
}
EUVIMED