Skip to content

1. Artificial Intelligence / AI

The root field: building systems that perform tasks normally requiring human intelligence. Its children are paradigms (how intelligence is achieved) and capability classes (what it does). The deep split is symbolic (explicit rules and logic, human-readable) vs statistical / learned (patterns extracted from data, opaque) — everything below inherits from that tension. "generative / predictive / multimodal / agentic" are not paradigms but behavior profiles layered on top of machine learning.

Children

  • symbolic AI — intelligence from explicit, human-authored structure:
  • logic systems
  • rule-based systems
  • expert systems
  • knowledge representation
  • statistical AI — intelligence from data and probability
  • machine learning — see Machine Learning
  • deep learning — see Deep Learning
  • generative AI — produces new content
  • predictive AI — forecasts / classifies
  • multimodal AI — spans text, image, audio, video
  • agentic AI — pursues goals via action; see Agents