Machine Translation: Context

In a scene from the feature film “Star Trek IV: The Voyage Home” Scotty talks to a computer without realizing that the user interface of the Macintosh was limited to a keyboard and a mouse. “How quaint” indeed.

John Searle's Chinese Room argument takes the Turing Test one deliberate step into obfuscation, by wrapping the original test for artificial intelligence between two layers of Machine Translation.

So what does it take to produce Machine Translation (MT)? Here is a word out of context, “wind.” What is the definition of the word? Is it a breeze, sounding like the word win? Or is a turning motion, sounding like the word wine? There is no way to tell because the word lacks context. So MT must figure out context to identify the correct definition of words, which is at the foundation of meaning.

Consider the word “right.” If we look at a dictionary, then we'll notice many different meanings for the word “right”. As humans we figure out which is the correct meaning by looking at the surrounding words, by establishing a context. This is my proposed solution for Machine Translation. If MT becomes successful, then the Chinese Room argument reverts back to the Turing Test, and the misdirection into Chinese is eliminated exposing the Chinese Room argument for a ploy to confuse the issue of artificial intelligence, rather than illuminate it.