thoughts on MathematicalInterfaceForLanguage
(github.com)
You're viewing a single comment thread. View all comments, or full comment thread.
Comments (2)
sorted by:
Actually it isn't. Not only that, but in many cases each sub domain it's used in (physics areas for example) people often create their own private language for some things, and use common math language for others. A big nasty example is quantum mechanics which is very difficult because so many QM nerds each create their own notation and language. A page of QM math literally is like trying to understand Sanskrit. It's completely alien.
Anyway, this MIL appears to be very naive so I suspect it's been written by a college student who's never taken a course in language theory. His writeup has some clueless concepts in it wherein he tries to apply the logic of geometry to the problem of understanding language. The world is analog and not digital, and it is probablistic. So binary geometric logic fails, for example in predictive situations.
"Will it rain tomorrow?"
I went to read his code and it is 11 years old, long before our modern era and its approaches.