December 3, 2025Open Access

Secrecy in AI: A Fragile Paradigm vs. the Open Internet Model

View Full Paper

Ask AI

Bookmark

Authors

Discussion

The analysis reveals how secrecy in AI models may hinder innovation, suggesting a transition to open standards is necessary.

Secrecy in AI can lead to narrative defense, making systems vulnerable and fragile, contrasting with open Internet paradigms.
Openness in the Internet allowed for peer review, collaboration, and innovation, elements often missing in AI today.
The paper discusses how a lack of transparency in AI architecture mirrors early financial industry bubbles with illusions of stability.
Emphasizing the need for foundational robustness and governance, the paper advocates for a shift back to shared standards and openness.

Guo Xiang-yu (2025) studied this question.

View Full Paper

Ask AI

Bookmark