Can Artificial Intelligence Answer Our Email?

Author

Adrian Bridgwater, Contributor

February 22, 2016

Many of us harbor a love hate relationship with email. We hate its Sisyphean tedium, but we love to stay on top of our world and our work objectives. Ultimately, many of us end becoming up a slave to our own inboxes. The question now then is… given the increasing sophistication seen in machine learning and Artificial Intelligence (AI), could intelligent automation answer our emails for us?

Developers to build-in more AI?

This rise in the deployment of automated machine intelligence technology is spiraling and, if popular, it could point software application developers towards putting a whole lot more AI into all our apps.

Google has already started to offer a degree of ‘robot email’ AI in its Google Inbox online extension for Gmail. The new service is called Smart Reply and it will help answer some of your emails for you. Inbox uses machine learning to recognize emails that need responses and to generate the natural language responses on the fly.

“Smart Reply suggests up to three responses based on the emails you get. For those emails that only need a quick response, it can take care of the thinking and save precious time spent typing. And for those emails that require a bit more thought, it gives you a jump start so you can respond right away,” writes Google software engineer Bálint Miklós on the firm’s own blog.

Google isn’t the only player trying to put more AI forward. LinkedIn is currently calculating how best to email users — it’s new Air Traffic Controller (ATC) platform for member communication including email, mobile and SMS uses a learning algorithm. For the LinkedIn, it appears, success comes in the form of fewer complaints.

“As a result [of this new technology], we are now sending 50% fewer emails and have reduced complaints by more than 65%,” said the firm’s Erica Lockheimer.

Facebook is using related machine learning technologies to help decipher the content of photos, object by object. They’re all at it, basically.

Not just the big boys/girls

It’s not just the industry titans playing this game; new startup Knowmail is also having a stab at the AI email market. This technology acts not instead of but alongside and inside of existing email clients. The software itself provides messaging services including prioritization, smart summary and ‘next best action’ on Outlook for Windows.

According to Knowmail, “The first step is helping companies tackle email overload, which could recover almost a million dollars a year in lost productivity per 1000 employees. But as we know, email is just the tip of the communications iceberg, so Knowmail aims to deploy its AI engine across the entire communications spectrum to ensure individuals receive the right messages at the right time on the right platform.”

Knowmail CEO Haim Senior explains that his firm’s concept was to present messages by importance. This means that Knowmail will learn your contacts, emails and habits and will prioritize things for you.
Inbox zero at last then?

Is this product the answer to all our woes? No of course it’s not. It’s AN answer in a field of many answers.

Some people don’t like email grouping into ‘threads’ (you can disable this in Gmail, but it’s still hard to get your Android device to do it), some people don’t like automation and wont agree with the walls that define Knowmail’s ‘next best action’ command. But some will — and that’s the problem.

The challenge that software application developers have when now building Artificial Intelligence into our apps is making that AI really act like a human.

In fairness to Knowmail, the so-called ‘state of mind’ filters are a clever attempt at doing just that. This function allows you to tell your email client that you are a) on the run or b) sitting down. This does indeed give the machine (your computer) more insight into how you are behaving.

Knowmail even shows an estimate of time it will take to complete the list of messages… and the system will ‘learn you’ and your typical behavior over time as a user. So we are indeed getting there i.e. we are teaching our computers how we humans behave.

Email is ripe for reinvention and this may just be part of the story… at least we’re on chapter one.

This article was written by Adrian Bridgwater from Forbes and was legally licensed through the NewsCred publisher network.


Comment this article

Great ! Thanks for your subscription !

You will soon receive the first Content Loop Newsletter