HTM, semantics, and fuzzy prediction

Creating a new thread, to avoid off-topic discussion.

If I understand your comment correctly, I think this is a misunderstanding of HTM. It does more than hard memory saves. It will also make fuzzy predictions based on past learnings of multiple contexts which share varying percentages of semantic similarities with a new novel context.

2 Likes

If you don’t understand how the fuzziness allows semantic generalization, this video helps:

4 Likes