I recently wrote an article about the contradictions I perceive in the current debate on Artificial General Intelligence. In that article I wrote: It’s beautifu

I recently wrote an article about the contradictions I perceive in the current debate on Artificial General Intelligence. In that article I wrote: It’s beautiful that we are able to distill a kind of inner structure of human knowledge by merely attempting to predict the next word in a sentence, based on the context. […] It’s as if there are certain arithmetic or algebraic laws of language that we don’t fully understand yet, but which make some sequences of words meaningful and human-like, while others are not. As soon as I wrote those lines, I knew I had to investigate those arithmetic or algebraic laws of language. My idea is quite simple. What happens when we combine vector embeddings (add, substract, multiply…) and what is the impact on the sentences we can get back after such a process ?