A compact taxonomy puts AI into seven core concepts from learning to logic
docmilanfar · x · 2026-07-28
A simple analogy between literature and AI: just as stories can be reduced to seven classic plot types, AI can be framed around seven core concepts:
- supervised learning
- unsupervised learning
- reinforcement learning
- representation learning
- search/optimization
- probabilistic/statistical inference
- symbolic reasoning and logic
It is more of a conceptual taxonomy than a technical claim, but it offers a compact way to think about the field.
More from AGI Musings
- Greg Conti says AI could erase authorship before expanding into wider human labor — wfithian · 2026-07-28
- A parent says AI made them realize it may be the last time they told their daughter to become a writer — wfithian · 2026-07-28
- HN Shah: intelligence should increasingly belong close to the person — msg · 2026-07-28
- Musk says the next frontier is a place with no training data — elonmusk · 2026-07-28
- A wry AGI take: maybe it just means knowing there is no solution — sull · 2026-07-28
- AI framed as a nuclear bomb, not a printing press, in debate over power concentration — iamtrask · 2026-07-28