Your Writing and AI

Does Rodin use your writing to train AI?

No. Rodin uses your writing to generate your intellectual fingerprint — a structured analysis of how you think that belongs to you. Your writing is not used to train AI models, is not shared with third parties, and can be permanently deleted at any time.

The distinction matters because AI companies have, in documented cases, trained on human-created content without consent or compensation. Rodin is built around the opposite premise: your writing generates something for you.

What AI companies have done with human writing

The scale of AI training on human-created content is not fully disclosed, but the documented cases establish the pattern. OpenAI was reported by the New York Times to have scraped over one million hours of YouTube videos to train GPT-4 — content created by individual people who received nothing in return. Google subsequently confirmed it uses YouTube content to train Gemini. Anthropic was found in federal court to have knowingly used over seven million pirated books.

If you have published writing anywhere on the public internet — essays, social media posts, forum comments, Substack newsletters — that writing has very likely contributed to training AI systems. The models became more capable; the writers were not compensated and were not asked.

How Rodin is different

Your thinking should work for you

Technologist Jaron Lanier has argued for what he calls data dignity: the idea that people should receive something in return for the data that AI systems are trained on. Your writing, notes, and accumulated thinking represent genuine intellectual labor. The current model extracts that value without returning it.

Rodin is not a compensation mechanism — it does not pay you for your writing. But it does something different from the standard model: it takes your writing and produces something that works for you. Your fingerprint surfaces what your notes already knew about how you think, and it finds you the people whose thinking resonates with yours. Your writing generates your intellectual identity rather than disappearing into someone else's model.

Common questions

Does Rodin use my writing to train AI models?

No. Your writing is used to generate your fingerprint and to support your profile over time. It is not used to train AI models — not Rodin's AI, not the underlying models it uses. You can delete your profile and all associated writing at any time via your management link.

Do AI companies train on my Obsidian vault or personal notes?

Not directly — unless your notes are published somewhere publicly accessible. AI companies train on content they can reach: public websites, social media posts, publicly shared documents, forum comments. Private Obsidian vaults stored locally on your machine are not accessible. But if you have published writing — essays, Substack posts, tweets, Reddit comments, YouTube transcripts — that writing is very likely in training sets.

Have AI companies used content without permission?

Yes, in documented cases. OpenAI was reported by the New York Times to have scraped over one million hours of YouTube videos to train GPT-4, which violated YouTube's terms of service at the time. Google later confirmed it uses YouTube content to train Gemini. Anthropic was found by a federal judge to have knowingly downloaded over seven million pirated books to train Claude, resulting in a settlement. These are the documented cases; the full scope of training data sourcing across the industry is not publicly disclosed.

Why does this matter to me as a writer or note-taker?

If you have published writing on the public internet — blog posts, social media, Substack, public forums — it has likely contributed to training AI systems that now compete with human writers. You received no compensation and no credit. The structural question this raises is whether there is a model in which your writing generates value for you rather than primarily for AI companies.

What is data dignity?

Data dignity is the idea, developed by technologist Jaron Lanier, that people should be compensated for the data that AI systems train on. Under data dignity, your writing would not be taken as raw material but would be treated as a contribution with economic value — something that generates a return for you. It is a proposed alternative to Universal Basic Income as a response to AI-driven displacement: instead of receiving state support as a passive recipient, people remain economically active as valued contributors to AI systems.

How is Rodin different from AI tools that use your content?

Rodin uses your writing to produce something for you — an intellectual fingerprint that belongs to you and connects you with people who think like you. It does not aggregate your writing into a training set. The relationship is the opposite of the standard AI model: instead of your writing making an AI system more capable, your writing makes your own intellectual identity more visible.

What does Rodin actually store?

Rodin stores a portion of your submitted writing to support your profile, enable fingerprint evolution over time, and power matching with other thinkers. Your writing is stored securely, used only for your own fingerprint and matching, and can be permanently deleted at any time. It is not shared with third parties, sold, or used in AI training.

Can I delete everything?

Yes. When you create a Rodin profile, you receive a management link with a unique token. Visit that link to access profile management, where you can permanently delete your profile and all associated data — writing, fingerprint, and cognitive signature — at any time.