AI Manifesto
AI can be a powerful tool. Used well, it can support research, writing, coding, and exploration. Used poorly, it can weaken accountability, reduce learning, and create risks around privacy, consent, and quality.
This manifesto defines how VSHN wants to approach the use of AI. Its goal is not to reject these tools, but to use them consciously, responsibly, and with clear human ownership. AI may assist our work, but it must not replace human judgment, responsibility, or care.
At its core, this manifesto reflects a basic rule of human interaction: if we want others to value our work, we must put real thought and care into it. AI can help reduce effort, but when it is used to bypass care and judgment, that usually shows - and the result feels low-value to the person receiving it.
This document is structured in three parts:
Core Principles, which define our general stance
Hard Boundaries, which define what we do not do
Practical Guidelines, which provide examples and day-to-day guidance for working with these tools.
Core Principles
- Responsibility remains with the human
AI does not reduce our accountability. Anyone using AI tools remains fully responsible for the quality, correctness, and consequences of the output.
- Judgement remains with the human
AI may assist human work, but it must not replace thought, authorship, review, or decision-making.
- Curiosity remains with the human
AI should support learning and exploration, not replace them. We use AI to deepen understanding, not to avoid it.
- Consent, privacy, and dignity must be respected
The input you provide must not violate privacy, ignore consent, or treat other people’s work and words as raw material without regard for their rights.
- AI use should be justified and proportionate
Because generative AI carries environmental, social, and labor costs, it should only be used where the value justifies the impact.
Hard Boundaries
We use only approved AI tools and follow all legal, compliance, and contractual restrictions
Here is a link to our compliance: <INSERT FUTURE LINK>
AI-generated work for production systems must be vetted by a human
Any AI-generated contributions to productive systems must be reviewed and verified by a human before it is shared, published, or merged. The responsible author remains accountable for the result and its mistakes.
In this context, “productive system” includes public-facing documentation, blogLow-effort posts, marketing content, etc.
We don’t shift the review burden
AI creates content extremely fast. With that in mind, please ensure you understand the content first before you ask others to review it. Review fatigue is real and has increased immensely with the onset of AI. Let’s not burden our people more.
AI is used to improve quality, not quantity
AI may be used to improve clarity, quality, and effectiveness, but not to increase the volume of output beyond what a human can reasonably understand, review, and process.
We don’t use any unsupervised AI tools acting publicly on behalf of VSHN
AI tools must not take any autonomous public actions on behalf of VSHN without human oversight and approval.
At VSHN, we do not use low-effort generated creative assets in official output
As low-effort we qualify art that has only been prompted without any further thought or modification to put the “humanity” into it. This includes:
Low-effort AI generated Pictures
Low-effort AI generated Music and Video content