​

Why it matters

AI terminology is inconsistent across research papers, documentation, and online content. Many concepts overlap, are misused, or lack structured relationships.

Mapping core AI concepts inside Geo will:

  • Standardize definitions

  • Clarify relationships between ideas

  • Improve search and discoverability

  • Reduce duplication across topics

  • Create a reusable ontology foundation for future AI bounties

This becomes the base layer of the AI knowledge graph.

What to publish

  • Topic entities for foundational AI concepts across areas such as machine learning, deep learning, neural networks, optimization, probability and statistics, reinforcement learning, natural language processing, computer vision, model training, model evaluation, and AI safety

  • Key properties per topic including name, clear and technically accurate definition, and at least one reputable source

  • Create meaningful relationships between related concepts to form a connected conceptual graph

Scope

Target 15 to 30 foundational AI topics.

Prioritize academically grounded and widely referenced concepts. Avoid buzzwords without clear technical meaning.

Potential sources

  • Technical encyclopedias and reference works

  • Peer-reviewed journal articles

  • Industry white papers and technical reports

  • AI research lab blogs and glossaries