How Machine Learning Can Improve Our Email Experience – Forbes


Around 2015 or 2016, Google began a strategic move, possibly one of the most important it has undertaken in its more than 20 years of existence, to incorporate machine learning and apply it to all its products. Its CEO, Sundar Pichai, saw machine learning as the ultimate disruptive technology, in the same way fire or electricity once were, and in a strong display of leadership, embarked on a mission to introduce the technology throughout the company, training all its employees in it.
A few years on, what can we observe? I think email, a tool we all use quite regularly, is an interesting example. What changes have we noticed over the last few years with Gmail, for example? For me, the most noticeable and practical difference is its anti-spam filter: like most people, I receive a lot of spam everyday, and although I still check my spam folder to see if something has been put there that shouldn’t, the truth is that these days, I check it superficially: frankly, I don’t even remember the last time I found something in that folder that wasn’t really spam. Nor do I receive spam in my main folders: for quite some time now, all spam goes straight to its folder, where I cast an eye over it and delete it without further ado.
Not so long ago, things were very different: spam messages would show up in my folders, I would need to label them as such, or I would have to check to make sure I didn’t delete something that shouldn’t be deleted. I still receive messages from assorted African princes, from lawyers notifying me of the death of a relative who has left me a cattle ranch in Argentina, advertisements for pills to maintain an erection for hours, along with notifications about bitcoin income in accounts I don’t have and in which I presumably have to identify myself to cash out… but I no longer see more than the title of the message, an instant before I hit the delete button.

I have also noticed a significant improvement in the time it takes for Gmail to understand that something is spam for me, even if it is not spam as such. After sending the same mail to the spam bin a couple of times, for example, press releases from sources I’m no longer interested in (the vast majority, actually), they stop appearing in my inbox, a learning process that used to take longer and that sometimes led me to create a rule for it. Now, if we also add the function that unsubscribes me from mailing lists that I never joined, I can’t even remember the last time I had to apply that rule. For quite some time now, spam is a problem that has practically disappeared, and to which I dedicate practically no time, despite receiving a ton of mail each day.

In 2018, Google introduced smart compose, which effectively autocompletes my emails, and in such a smart way that it borders on the worrying, prompting me to wonder how Google can know what I was thinking of saying to a given person. A feature I never thought would ever become practical, and yet I find myself making use of it on more and more occasions. For some time now, the checker no longer just notifies me of typos, but has become a pretty good grammar checker — which is very useful when you frequently write in your second language.
Gmail has been my main mail manager for a long time, long before, for example, I was allowed to use it for work purposes. I’ve had my email archive in Gmail since 2004, which is really impressive when I think about how few emails I delete and how easy it is to locate an old one when I need it. Google’s use of machine learning means I will stick with Gmail for the foreseeable future. When I think about Gmail’s annoying habit of displaying ads when I use it from my smartphone (which I assure you I will never click on, even if they advertise something I’ve been wanting all my life), I would rather pay for them to be removed, rather than using another email service
Machine learning can contribute hugely to our use a particular product. But above all, it can build competitive advantages that generate loyalty, and make them realize, as has happened to me, that switching to rival product makes no sense. Applying Google’s Machine Learning to Gmail is a good test of that, and a prove that a strong leadership strategy, like the one undertaken by Sundar Pichai five years ago placing his bets on machine learning, can yield very interesting rewards and allow you to greatly differentiate your products from those of competitors. 
Does this match your experience, or am I simply very lucky (or perhaps utterly predictable)?