Pages

Google Will Utilize AI And Machine Learning Algorithms To Fine Tune Gmail Suggestions

Google's setting up Gmail for an upgrade, as it employs ML models to help users with better search suggestions.

Google has been tampering with, and soon after disposing of, new technology and updates since its inception. We have the Google Graveyard as a testament to such haphazard progress. However, let's be real: the tech giant knows when it's onto a good thing, and Gmail is that good thing. The email-based service is one of Google's trademark platforms and continues to flourish even as emails are considered more and more restricting. No one types out an email to converse when they can use WhatsApp or Instagram. However, rarely do people utilize the same platforms to send out CVs or letters of recommendation: even if they could, it would be considered informal and inappropriate.

That is the sort of environment Google has managed to cultivate with Gmail. It's what Hotmail and AOL chat rooms could never really muster up; Gmail is a brand that resonates with professionalism and the company leans into it as well. This brings us to today's feature of interest, and how Google will utilize it to the email service's full effect.

ML or machine learning algorithms have evolved from being a novel, rarely used concept that evoked "oohs" and "aahs", into a tool that's almost regularly utilized by every social media platform with skin in the game. Honestly, they fit the social media market quite well: ML algorithms rely on using tons of data to generate whatever they're ordered. Automated messages can appear more natural if AI spends its time looking at and attempting to emulate examples of normal speech. With social media platforms (I'm willing to loosen the definition enough to include email platforms) providing a near-endless well of such information to draw from, it's open season for ML programs.

Google intends on utilizing ML models to help AI-based suggestions grow with the writer. To be clearer, Gmail's word suggestions for users will come to reflect what the user is aiming for based on previous interactions. This way, a consumer's suggested vernacular can essentially come forward as a more elegant version of what they'd employ in day-to-day life.

These suggestions will also be heavily utilized in helping users sift through folders and such, attempting to look for prior emails or other similar content. ML models can learn from the keywords that a user utilizes in looking up content in the past, and help narrow down searches in the future.

Google will be rolling out these new updated ML models for user testing across Android audiences.


Read next: Google Is On A Clean-Up Mission To Remove One-Star Scam Reviews That Tarnish Small Brands’ Reputation

No comments: