LinkedIn boosts its messaging with smart replies, pre-written, AI based interactions

LinkedIn — the Microsoft-owned platform for those who want to network with professional contacts and advance their own careers — has been in the middle of a long-term makeover of its social tools, as it looks to drive more usage. Today comes the latest chapter in that story: the site is unveiling a new smart reply feature in its messaging app, which gives users prompts with different phrases to use while they are chatting to keep the conversation flowing.

The feature is launching in English first in LinkedIn’s mobile app and on desktop. LinkedIn says that it plans to roll it out to more languages sometime in the future. And users can opt out of the smart reply feature in their settings.

Smart replies may sound familiar to you for a couple of reasons. The first of these is that LinkedIn itself has been trying out a version of suggested replies since January of this year, and was actually already talking about its plans on this front months before that.

The key difference in today’s news is that the company is now using more AI tools like machine learning and more sophisticated natural language processing to be able to understand the gist of a conversation and how to help keep it going.

Arpit Dhariwal, a senior product manager at LinkedIn, said that the company is working on updates to smart replies that will continue to make it more personalised.

The second reason why smart replies might sound familiar is that the concept of these quickly becoming the rule rather than the exception. Earlier this year, Google expanded its own version of the feature, also imaginatively named Smart Reply, which had first made its debut in its AI-infused Inbox app, into its much more ubiquitous Gmail app.

There are multiple reasons behind why LinkedIn, Google and others are working on ways of making it easier and faster to reply to messages.

Perhaps the biggest of these is that they help get more people using their messaging services. Messaging apps (and that includes email messaging) have become the most valuable property on smartphones these days as a hugely popular way for people to communicate to each other, encroaching on phone calls and other native phone features.

LinkedIn itself realised this years ago and has been trying to improve its messaging experience ever since.

But although there are some definite demons in the world of touchscreen typing, in general it can be a pain to compose messages on smartphones and tablets (and for some it’s a pain on regular keyboards, too). It’s all the more so when the messages are mundane interactions.

Predictive phrases and words aim to kill those two birds with one stone, by making the most obvious/typical replies into one-tap buttons, phrases that you can use as a starting point that you can edit after inserting them, or just leave them as they are.

The other area that’s interesting here is how LinkedIn is trying to do a lot more with artificial intelligence. As AI represents the new wave of technology, LinkedIn is making sure that it stays in the loop to use it where it can in products it builds.

That’s not entirely new of course: AI tools like machine learning have been used by LinkedIn in its backend and select products some time now, and I’m guessing we will be seeing more of that to come.