Fact Claims - is your data trustworthy?

Our goal is to enrich the world with data that is relevant, licensed, and trustable at planetary-scale.

Recently, mediocre web content has experienced exponential growth - bulk-automated by generative AI.

If our future AI are trained on this - we can expect increasingly lower-quality halluciations.

Ensuring the integrity and trustworthiness of our AI systems is important:

• To learn more: start with our white paper then learn the basics.

• For the more technical: read the RFC and our raw facts.

For content owners

• Publishing your own fact claims allows intelligent systems to discover your expertise, offerings and license terms.

For developers, researchers and crawlers

• Discover, learn and engage with rich, contextual, quality data to improve the signal/noise ratio for your analysis, RAG and fine-tuning.

Online Examples:

• fact.claims website
• claims creation wizard
• fact map demo

Fact Architecture:

• Valid Facts: For a fact to be considered valid, it needs to be associated with a specific action or entity and generated by a particular process.

• Traceable Facts: These facts can be traced back to their source through connections established during their creation or association.

• Grounded Facts: Grounded facts are securely stored in an unchangeable document, ensuring their reliability and verification.

• Published Facts: Published facts are online for reference, they need not be immutable so may be real time.

• Notarized Facts: Notarized facts with a trusted chain of custody, with provenance traced back to a digital trustee.

Explore Further:

• Trustworthy AI

  Learn about building AI based on facts, trust and integrity.

• White Paper

  Access our comprehensive white paper on trustworthy AI and fact claims.

• RFC v0.4 DRAFT

  Review the latest version of the Request for Comments (RFC) for our project.

• How-to Guides

  Dive into practical guides for managing and publishing fact claims.

• Learn the basics

  Get started with the basics of fact claims and their importance.

• Write your first facts claims

  Step-by-step instructions to publish your own fact claims.

• Curating and crawling

  Learn about curating and crawling data to ensure trustworthy fact claims.

• Querying facts with SPARQL

  Learn how to query with SPARQL to retrieve the facts you need.

• SEO for AI

  Learn how to repurpose existing SEO rich snippets as fact.claims.