Article Details

Abstract

Addressing the lack of systematic analysis on the inherent developmental patterns of vertical large model ecosystems in existing research, this study, grounded in technological affordance theory, constructs a three-dimensional analytical framework encompassing core capability affordance, access and integration affordance, and collaborative evolution affordance. Through a cross-case analysis of sectors such as healthcare, finance, and autonomous driving, the study reveals how the design of technological systems, as a key moderating variable, shapes the action possibilities of different actors and fosters differentiated logics of value creation and distribution within vertical large model ecosystems. The research identifies three common mechanisms: the value consensus mechanism based on trustworthy affordance, the innovation diffusion mechanism driven by modular affordance, and the ecosystem lock-in and counter-lock-in mechanisms triggered by imbalanced affordance configurations. The findings indicate that competition in vertical large models is essentially a competition of technological affordance configuration models, offering a theoretical perspective and practical framework for understanding and governing vertical large model ecosystems.

Keywords

vertical Large Model; ecosystem; technological affordance; value logic; mechanism

References

1. [1] Li Xinyao, Li Jingjing, Zhu Lei, etc A review of efficient transfer learning algorithms for resource constrained large models [J]. Journal of Computer Science, 2024, 47 (11): 2491-2521
2. [2] Jiang Nan, Suo Hansheng, Tang Ye, etc Research and Application of Multi Domain Adaptive Expert Separation Large Model Framework in Petrochemical Industry [J]. Journal of Chemical Engineering, 2026, 77 (7): 4018-4029
3. [3] Cao Rongrong, Liu Lin, Yu Yandong, etc A review of research on large language models integrating knowledge graphs [J]. Computer Application Research, 2025, 42 (8): 2255-2266
4. [4] Yao Yi, Chen Chaoyang, Du Xiaoming, etc Overview of Multimodal Knowledge Graph Construction Technology and Its Application in Military Field [J]. Computer Engineering and Applications, 2024, 60 (22): 18-37
5. [5] Tang Huajun, Wu Wenbin, Yu Qiangyi, etc Construction of a lightweight large-scale model for remote sensing monitoring of cultivated land based on morphological fingerprint features [J]. Chinese Journal of Agricultural Sciences, 2026, 59 (1): 78-89
6. [6] Chen Siru, Shu Yuanchao A review of edge deployment and inference acceleration techniques for multimodal large models [J]. Journal of Zhejiang University (Engineering Edition), 2026, 60 (4): 723-737
7. [7] Wang Bo, Niu Chaowen From ChatGPT to GovGPT: Building a Government Service Ecosystem Driven by Generative Artificial Intelligence [J]. E-Government, 2023 (9): 25-38
8. [8] Tang Luyuan, Xie Shiyao, Xu Yuan Research on the Construction of Innovation Ecosystem and Risk Identification in Large Model Industry [J]. Scientific Research, 2025, 43 (9): 1972-1981
9. [9] Liu Liu, Zhang Xin, Zhang Qiqi, etc Research on Ethical Risk Identification, Governance, and Prospective Analysis of Large Models in the Medical Field [J]. China Engineering Science, 2025, 27 (6): 54-67
10. [10] Yang Qingfeng The operating mechanism and development strategy of generative economy [J]. Gansu Social Sciences, 2025 (1): 196-206
11. [11] Zhang Hui, Chen Zhengwen, Zeng Xiong Analysis of the Innovation Mechanism and Its Effects of Deep Search Company: Taking the DeepSeek-R1 Large Model as an Example [J]. Scientific Research, 2026, 44 (4): 770-779
12. [12] Xie Weihong, Zeng Simin, Peng Tiepeng, etc Technical availability: conceptual connotation, theoretical framework, and prospects [J]. Science and Technology Management Research, 2022, 42 (5): 210-218
13. [13] Yao Junchen, Xia Yining, Zhang Hongzhong Access, Switching, and Nomadism: A Study on User Multi Platform Usage Behavior in the Context of Large Models and Its Governance Implications [J]. Journal of China University of Political Science and Law, 2026 (1): 152-168
14. [14] Chen Zhi, Chen Hao Empowering utility and risk perspective of ChatGPT from the perspective of availability [J]. Intelligence Journal, 2023, 42 (7): 131-139
15. [15] Chen Siyu, Wang Qizhen, Zhang Hongzhong The Art of Questioning: Analysis of the Availability of Large Model Hint Engineering [J]. Modern Publishing, 2025 (4): 61-72
16. [16] Pan Hongliang, Gu Hang, Ye Lu The Impact of Artificial Intelligence Industry Availability on Innovation Quality of Manufacturing Enterprises [J/OL]. Research Management, 2025:1-17 https://link. cnki.net/urlid/11.1567.G3.20251212.1314.002.

Funding

Shanghai High level Institution Construction and Operation Plan "Soft Science Research" Project (25692106300)

About the authors

Wang Zhenming (1981-), male, Han ethnicity, from Yushu, Jilin, senior engineer, doctoral degree, research interests include artificial intelligence, the Internet of Things and big data; Ma Na (1985-), female, Han ethnicity, from Changchun, Jilin, engineer, master's degree, research interests include AI algorithm research; Jiang Shan (1991-), male, Han ethnicity, from Xinyang, Henan, engineer, master's degree, research interests include signal processing and computer vision; Corresponding author: Wu Bo (1980-), male, Han ethnicity, from Zibo, Shandong, senior engineer (full grade), doctoral degree, research interests include the Internet of Things and the Industrial Internet and industrial intelligence.