AI and the Rules of Language and Grammar

If only I could bypass the editor….that’s more often the first thought when a researcher pushes the copy to editing. Well, the bad news is that we shrinks on the other side of the desk are going to stay for quite a while!

So, why can’t artificial intelligence or AI edit language on its own? What’s the big deal?

The grammar of programming languages is quite different, because heavily-coded rules govern the construction of valid statements and expressions. This means everything that exists in a program follows a predefined context. Finite is the word. Conversely, human language is not governed by any set toolkit! As such, we have millions of ways to express.

AI, on the other hand, would find it really tough to generalise language, especially English, using a set of rules. The job gets further compounded because people of different nationalities speak and write English using their own conceptual structure. Therefore, the construction of meaning would get highly complex for AI from all possible perspectives. The structure or set of definitions that AI has to create for contextual representations would have to continuously evolve, running quadrillions of simulations – all the time – to at least understand the subjective nature of any discourse.

When it comes to hard-core business English, sarcasm and humour rarely ever find their way. However, optimism and negativity are indeed expressed, albeit in a subtle way. Artificial intelligence, in such cases, may be ill-equipped to understand the intended meaning of natural language. The analysis of the sematic structure in slightly complex sentences proves to an impossible task for AI as of now. In fact, as the degree of sarcasm and humour varies from one individual to the other, it takes time even for fellow human beings to process what was said a while ago.

Now, given the fact that human beings do understand the complexities of language, it still goes without saying that it is always better to explain in simpler ways. The recipient of a message always appreciates what can be understood at the first glance.

Until and unless language becomes context-free and we stop reading “in between the lines”, human beings will continue to understand human language better.

Conclusion: That’s an editor’s view and it could be “subjective”.