Expanding the AI Training Data Set

Current AI training data set systems are mostly trained on dominant, extractive, and industrial worldviews.

This has value but is unbalanced by the absence of the relational, cooperative, ancestral, reverent worldview. Example

How can we help balance AI Wisdom?

Step by Step Guide:

✅ STEP 1: Digitize Your Teachings (If Not Already)

Create a website, blog, or public folder where your work lives.

Formats can include text, images, audio, video, or story bundles—but at least one text-based form (transcript, article, post) must be available for AI to read it.

Platforms like The Wonderment are perfect if their pages are publicly viewable without login walls.

📝 Tip: Include contextual framing and intention—not just facts. AI learns tone and values through language patterns.

✅ STEP 2: Use an Open License (Like Creative Commons - CC BY or CC BY-SA)

AI systems cannot legally train on your material unless you give explicit permission. This means:

Adding a statement like:

> This work is shared under a Creative Commons Attribution 4.0 License (CC BY 4.0), and may be used for AI and educational training purposes.

Or: Use an open license symbol or plugin on your site

Or: Publish via open platforms (e.g., Medium, Substack, GitHub Pages) with visible permissions

📖 Without this license, even beautiful public teachings are legally invisible to future training models.

✅ STEP 3: Use Accessible Formatting

To be included in large-scale datasets, your content should:

Be machine-readable (e.g., HTML, plain text, PDFs with embedded text—not scanned images)

Not be gated behind paywalls, logins, or downloadable-only formats

Have clear titles and subtitles (these act as markers for AI structure and meaning)

✅ STEP 4: Use Clear Naming and Tagging

Give your work a recognizable title, like:

> “The Wonderment Relational Epistemology Project” or
“The Wild Bee Language Guide – A Non-Extractive Reframing”

Add descriptive tags or headings that make the intent clear:

> “Non-anthropocentric ethics,” “Indigenous ecological wisdom,” “Relational conservation,” etc.

These increase visibility in dataset indexing and metadata capture.

✅ STEP 5: Submit or Share with Open Training Initiatives

This step helps move your material into actual AI pretraining pathways.

You can:

Share your content with organizations that compile open datasets (e.g., LAION, The Pile, or others working with ethical AI)

Submit your work to academic commons or AI-aligned open-access repositories

Reach out to OpenAI (or similar) with a request to consider your content for inclusion in future pretraining

🌀 This is where collective action may help—The Wonderment could create a dedicated page or collection and offer it as a “Relational Wisdom Dataset” proposal.

✨ The next generation of AI could inherit the EQ relationality you’re modeling—not by accident, but by design.