An open standard to signal human effort in AI-assisted content.
The Problem • The Standard • How to Use • Badge Generator • Platform Guides
You're scrolling through your feed. The prose is clean. The structure is solid. But your brain already made a decision: probably AI-generated, probably not worth my time.
Meanwhile, someone who spent hours researching, developing original ideas, and shaping a vision used AI to help write it. They get the same skeptical glance as content generated in 90 seconds.
To readers, these look the same.
No signal. No context. No way to know.
Time is the only currency that never comes back.
When someone spends hours developing original ideas and shares that with the world, they're giving something they can never recover. That deserves to be recognized. Not celebrated. Not rewarded with special treatment. Just seen.
Right now, follower count is a proxy for trust. Fame determines who gets read. That's not fair. And it's not accurate.
HUMI changes that. It gives every creator, regardless of audience size, a way to signal the effort behind what they made.
HUMI: Human Index
Four levels. Simple. No judgment, just clarity.
HUMI v1 focuses on written content: articles, newsletters, documentation, blog posts. The principle extends to other formats, but we start here.
|
No AI involved. The author created this entirely themselves. No generative AI was used in writing or production. Using AI for research doesn't affect this. Only AI involvement in creating the final content. |
|
Human created the thesis. AI helped express it. The author developed the original ideas, framework, or argument. AI assisted with writing and refinement. The thinking is human. The expression had help. |
|
Human steered. AI contributed to substance. Human decided. The author guided the content through real decisions. AI proposed ideas or structure; human selected and shaped the outcome. More than a prompt. Real collaboration. |
|
AI drafted. Human verified and owns the outcome. The author provided direction and reviewed for quality. AI generated the content. Fast, practical, accountable. |
A HUMI 3 product update doesn't need hours of original research. A HUMI 0 essay represents a choice to create without AI.
The point isn't judgment. It's clarity.
HUMI is self-reported. There's no verification mechanism, and that's intentional. The badge is a conversation starter, not an audit.
Think of HUMI as a chain of trust. A HUMI badge from a stranger means little on its own. But from someone you kind-of trust? It solidifies that trust. HUMI amplifies existing reputation—it doesn't create it from scratch.
Mislabeling is reputationally costly. If you claim HUMI 1 and your readers sense slop, you've lost more than you gained. The value is in honest signaling.
HUMI requires human accountability. If no one is willing to review it, refine it, and stand behind it, it doesn't get a HUMI badge.
Content with zero human accountability is outside the standard.
Put HUMI at the top. Above the title, or immediately after it.
Never at the bottom. By the time readers reach the bottom, they've already made their trust judgment.
Generate badges in seconds:
- Choose your level (0-3)
- Pick a theme (including minimal "Editorial" style for clean blogs)
- Export as HTML, Markdown, plain text, or PNG/SVG
- Copy once, paste everywhere
WordPress / Ghost / HTML Blogs
Use the HTML embed from the generator. Paste directly into your post editor (HTML mode).
<a href="https://github.com/fscheps/humi" style="display:inline-flex;align-items:center;gap:10px;padding:10px 16px;background:#1565C0;border-radius:8px;text-decoration:none;font-family:-apple-system,sans-serif;" target="_blank">
<span style="font-size:13px;font-weight:700;color:white;">HUMI 1</span>
<span style="width:1px;height:16px;background:rgba(255,255,255,0.3);"></span>
<span style="font-size:13px;font-weight:500;color:white;">Human Envisioned</span>
</a>For minimal blogs, use the Editorial theme for a subtle, text-style badge that blends with your design.
LinkedIn / X (Twitter) / Social Media
These platforms don't support HTML. Use:
Option 1: PNG Badge Download from the generator and upload as an image.
Option 2: Text Badge
HUMI 1 · Human Envisioned
Human created the thesis. AI helped express it.
Learn more: github.com/fscheps/humi
---
[Your content here]
GitHub / Reddit / Markdown
**HUMI 1** · Human Envisioned [↗](https://github.com/fscheps/humi)Or use shields.io badges:
[](https://github.com/fscheps/humi)
Medium / Substack
Use the PNG image badge from the generator at the top of your article. The Editorial theme works well for clean newsletter aesthetics.
Or add a text note:
HUMI 1 · Human Envisioned
Human created the thesis. AI helped express it.
| Level | Name | Color | Meaning |
|---|---|---|---|
| HUMI 0 | Human Created | #2E7D32 |
No AI involved |
| HUMI 1 | Human Envisioned | #1565C0 |
Human created the thesis. AI helped express it. |
| HUMI 2 | Human Directed | #00838F |
Human steered. AI contributed. Human decided. |
| HUMI 3 | Human Reviewed | #607D8B |
AI drafted. Human verified and owns it. |
Is HUMI 3 "bad"?
No. HUMI 3 is honest and efficient. A product changelog doesn't need hours of original research. The point is clarity, not judgment.
Is HUMI anti-AI?
No. HUMI recognizes that AI is a tool. The standard exists to signal effort, not to judge methods. All four levels are legitimate.
How do I choose between HUMI 1 and HUMI 2?
Use the Litmus Test: If you removed the AI, would the primary argument remain exactly the same?
- Yes → HUMI 1 (you created the thesis; AI helped express it)
- No → HUMI 2 (AI contributed to the substance; you steered and decided)
Another way to think about it:
- HUMI 1: You developed the framework or argument first, then AI helped write or refine it.
- HUMI 2: You prompted, AI proposed structure or ideas, you selected and shaped the outcome.
Examples:
- HUMI 1: Original framework + AI copyedit and rewrites
- HUMI 2: AI proposes alternative models or analogies; you pick, combine, and reshape
When in doubt, be conservative.
What if I'm not sure which level fits?
Ask yourself:
- Did I develop original ideas or frameworks? → HUMI 1
- Did I actively direct and shape through real decisions? → HUMI 2
- Did I mainly prompt and review? → HUMI 3
- No AI at all? → HUMI 0
When in doubt, be conservative.
Does using AI for research affect my level?
No. HUMI measures AI involvement in creating the content, not researching it. Using Perplexity or Claude for research while writing everything yourself is still HUMI 0.
Can I use HUMI for non-written content?
HUMI v1 focuses on written content. The principle extends to other formats (video, audio, images, code), but we're starting with text. Future versions may provide format-specific guidance.
Can I change my HUMI level after edits?
Yes. If you substantially revise content, you can update the badge to reflect the new balance of effort.
What stops people from lying about their level?
Nothing technical. HUMI is self-reported and trust-based.
But think of it as a chain of trust: a HUMI badge from a stranger means little. From someone you kind-of trust, it solidifies that trust. HUMI amplifies existing reputation—it doesn't create it from scratch.
Mislabeling is reputationally costly. If you claim HUMI 1 and your readers sense slop, you've lost more than you gained. The badge is a signal of good faith, not a certification.
HUMI only works if people use it.
As a creator:
- Start using HUMI on your content
- Be honest about which level fits
- Place it at the top, not the bottom
As a reader:
- Notice when creators use HUMI
- Appreciate the transparency
As an advocate:
- ⭐ Star this repo
- Share with your community
- Help build the norm
CC0 1.0 Universal (Public Domain)
Use it, adapt it, build on it. No attribution required.
Created with ❤️ in 🇨🇭 Switzerland by Fernando Scheps
