This diagram shows an effective determination in real time of possibly meaningful phrases that avoids the need for expensive power consumption and also splits the meaning of language from the knowledge.
This video was a short extract where I discuss the process from linguistic framework Role and Reference Grammar that we’ve implemented that converts language (words and phrases) to its meaning.
“It’s been the Holy Grail for 70 years.”
On the left side is the consolidation set (CS) that captures the possible syntactic elements and on the right is the language independent meaning representation, a sematic set (SS), that is validated by things like selectional restrictions.
This is actually a brief explanation of a patent that I was recently granted that allows an effective determination in real time of possibly meaningful phrases that avoids the need for expensive power consumption and also splits the meaning of language from the knowledge.
My latest book includes a roadmap enabled by this kind of model in which language use by machines becomes ubiquitous within 5 years in any language. We can do a lot of useful things without knowledge in many use-cases.
We can take advantage of developments in cognitive science like this one to rapidly improve AI and facilitate undreamed-of applications. All we need to do is to start.
Share this post