Novamente Research Papers and Source Notes
Find source notes for historical Novamente and OpenCog references including AGI architecture papers and related research pointers.
Archive
Historical Novamente Examples and Demos
Context for old Novamente demo, example, video, PetBrain, NLP, fetch, and virtual pet paths with modern reliability lessons.
This page catches old example and demo links from the historical Novamente/OpenCog period. It does not host or claim ownership of the original demos; it provides context and points readers toward source notes when available.
Important source paths this page replaces
/example//example/nlp.html/example/fetch.php/example/grab ball.html/zeptopetz/
Old NLP demo links
Old NLP example links usually came from the period when Novamente and OpenCog-related projects were trying to demonstrate interactive behavior, language handling, or agent-like loops. This page does not recreate those demos. It routes the intent to historical context and to modern verification patterns.
Old fetch demo links
Fetch-style demo references are best read as examples of an agent loop: observe, choose an action, execute it, and receive feedback. A current workflow should make each of those steps inspectable.
Old grab-ball demo links
Virtual action demos can be useful for explaining behavior, but a one-off behavior is not the same as a reliable workflow. Production systems need test fixtures, traces, stop rules, and rollback.
ZeptoPetz and virtual pet intent
Some old links pointed to virtual pet or PetBrain-style examples. The useful modern lesson is not that old demos should be copied directly. It is that agent behavior needs an observable loop: perception, state, action, feedback, and review.
Modern reliability lesson
A demo can show a behavior once. A production workflow needs evidence that the behavior is bounded, repeatable, recoverable, and reviewable.
Use this checklist when translating a demo into a workflow:
- What did the agent observe?
- What state did it store?
- What action was allowed?
- What evidence was logged?
- What feedback changed future behavior?
- Who could stop or correct it?
Related modern pages include Build a Research Agent, Build a Customer Support Knowledge Bot, Agent Observability Guide, and Agent Risk Scorecard.