Once models / inference engines are performant enough to run on consumer hardware without hogging all the resources of a machine -- sure! Embedding this into th

Once models / inference engines are performant enough to run on consumer hardware without hogging all the resources of a machine -- sure! Embedding this into the core of a linux system sounds very interesting. A shell interaction could be had in nerd gobbledygook or plain human language.