Remember how much you hated your English teacher's seemingly angry red pen all through school? In an age where very few people learned to write correctly in the first place due to growing up with smartphones that did it for them, Weird Al Yankovic's "Word Crimes" song is no longer just a funny anecdote.
Workplace polls over the years have shown employers expressing increased annoyance with the grammatical shortcomings plaguing the workforce across the board.
Gmail is bravely answering this common problem and making grammar correction friendlier with a new AI-based grammar tool. It will make sure you never send another work email that makes you (or your boss) wonder how you passed 7th-grade composition.
A big step up from foundational approaches to editing software, this new program brings machine learning to the forefront by showcasing your written expression at its best possible angle.
What's new and how did we get here
If you have used Google Docs anytime after February of this year, you have seen some rendition of this new software. The program offers inline spelling correction and contextual suggestions as you type. Gmail engineers worked closely with linguists bearing areas of expertise from computational analytics to sociology.
These linguists worked in panels of three to pore over thousands of grammar samples, which they then had to come to a consensus on regarding correct grammatical patterns that would be placed into the algorithm.
Semantics and sentence structure standards were put in place for over 40 languages and counting. The overall aim of these G Suite improvements is to use automated machine translation at a level that goes beyond the functions of digital dictionaries.
Why this is different from and better than basic spell check
Grammar is a significantly more complicated and elusive beast than spelling. While there may be one or two ways to spell any given word, there is still a clear right and wrong in that arena. Grammar, by contrast, is largely based on a series of intangible components, such as intended meaning, tone, inflection, and contextual relevance.
We've all had the cringe-worthy experience of hitting send on an email that suddenly looks grotesquely wrong in a sentence or two that may now inadvertently offend or misinform someone, and email is particularly conducive to making language uniquely dehumanized and thus, much more likely to be misunderstood.
Ultimately, mistakes in professional emails lead to wasted time, and wasted time leads to wasted money - something no one is in business for. Gmail hopes to help eliminate these productivity problems by combining billions of high-quality web-sourced words with the specialized skills of actual human beings trained in language.
While some may argue that a machine-based tool attempting to fight a problem that is likewise generated out of overuse of machines is counterintuitive - we say a little "hair of the dog" is sometimes just what the doctor ordered.