A forum post on zhichai.net reviews The Disclosure Penalty (2026.05), a study at the intersection of psychology and AI, exploring why people instinctively devalue creative work once told it was AI-generated.
Key points
- Blind tests vs. disclosure: When AI-generated poems, paintings, or music are mixed with human works and judged blind, participants often cannot tell the difference and may rate AI works highly. But once told "this one is AI-written," scores drop sharply — described as a forced distortion of aesthetic perception by prior beliefs.
- The disclosure penalty: Based on 16 comparative experiments, the study confirms that disclosure alone degrades ratings — not because the work changed, but because the brain applies a "defense filter" that strips the work of perceived emotional value, creative intent, and "soul."
- Empathy with the creator: Humans欣赏 art not just as pixels and words but by empathizing with the agent behind it. Knowing the "creator" is matrix multiplication collapses that empathetic connection.
- Aesthetic value is partly social: Artistic value exists half in the work's structure and half in the audience's sociological imagination about the creator's motives.
- In the AI era, the bottleneck is no longer technical quality but humans' psychological rejection of "silicon-based creativity."
- When marketing AI content, don't assume "quality speaks for itself" — consider the psychological cost of transparency.
- A business model built on "perfect deception" faces a severe trust collapse once AI authorship is disclosed; the challenge is helping audiences appreciate beauty that isn't backed by human suffering.