Running a final set of tests on the new version of my submind architecture.

It's now a recursive architecture (so subminds have their own subminds) and questions are broken down into multiple sub questions, which are then distributed to the most appropriate subminds.

Quantum Economist

Engineering Generosity: building abundant systems

Builder of 🧠 

Future First Creator to Win a Nobel

Impossible is an opinion

Then, when a submind has no available subminds, it simply performs a research step to try to answer the questions it came up with, compiles that into an answer for the question it received, and propagates it back up the chain until the original asker gets an answer.

That creates a really cool way to share knowledge without compromising personal data at all.

The cool part is that each submind has a current knowledge state that updates every time it figures out an answer. When each submind is only part of a single larger mind, that doesn't matter too much, but when it is part of many superminds, it learns from all questions it gets.