I seemed to have solved the problem of machines learning like people, I’ll place the source code for study below, but the main problem is how that information is stored for later use. In this language learning app, it takes a snapshot of the old file, and then appends new information to the file. So you can keep adding new words. And have that system throw a random word and defintiion at you. But the problem quickly becomes how many language instances can you store:
This is simple framework I’ve found to efficiently store information, the problem become a matter of that fact that this can quickly spiral into infinity. The best computers that we have are only able to store so much information.
The diagraph is based on the Nicola Tesla mod 9 concept, and taken further.
Source code:
https://github.com/LWFlouisa/DuckLingo/blob/master/README.md
Update: The other problem is, and why I dislike Decision Trees, is they assume everything operates is dependant on other things, and not operating as its own intelligence.
In practice, you need a brain that can operate multiple independant function, or you’ll have silly scenarios like:
I turn my brain on. While I read a book, I kick a door down. Machines would need each function to operate largely co-independantly, otherwise you might end up causing unintentional damage to your robot.