Abolfazl Sabramiz; Mehdi Zakeri; Roohollah Haghshens
Abstract
When we ask if Strong AI is possible we ask, in fact, if AI can lead to the level in which most basic human intelligence features achieved. In this research we ask this question more ...
Read More
When we ask if Strong AI is possible we ask, in fact, if AI can lead to the level in which most basic human intelligence features achieved. In this research we ask this question more specifically about understanding language. There are some philosophical challenges in the way of modeling linguistic understanding for AI Machines. Based on Frame Problem, i.e. updating beliefs based on new beliefs and behaviors, is raised as a frame problem in the literature of artificial intelligence as well as philosophy (in the field of epistemology), we show these challenges as: first, in a specific situation how we can distinguish relevant features for understanding a bit of language from irrelevant ones? second, if we understand a bit of language by inference from our previous knowledge about the world and previous linguistic knowledge, how we can understand a bit of language immediately? third, how we can understand polysemous and ambiguous words?
