Chatbots


Nathan Baranowski

June 28 2018

Whether AI truly is AI or just a well named marketing ploy is a topic for another article, but no-one can deny the use potential for ‘smart’ technology at the touch of a button. How are they going to be used in the future?

Chatbots are everywhere today – and not just in the news reports that pop up on your smartphone. In simple terms, chatbots are assistants.

They are, reportedly, one of the more visible ways that AI (Artificial Intelligence) is becoming integrated in our day to day lives. Whether AI truly is AI or just a well named marketing ploy is a topic for another article, but no-one can deny the use potential for ‘smart’ technology at the touch of a button.

Before we go into the nuances of what chatbots can do, let’s take a look at the niche they’re designed to fill.

Personal:

For personal devices, as with Google Assistant’s newest chatbot Duplex, the chatbot is designed to imitate life. It was demonstrably able to book a hair appointment for the user, speaking to a representative of the business and discussing availability, among other requirements.

This technology caused some backlash, however. While understandable, we’re still in early stages. The debate around dehumanising bots has yet to happen… right now we’re still working on demonstrating real-life applications.

In short, chatbots in personal lives are designed as an extension of the calendar – a sort of part-time personal assistant, a smarter version of your smart phone.

Business:

In business, however, we see a more holistic application. Instead of just booking appointments, chatbots can be used for customer service. By 2020, Gartner now predicts that 85% of customer interactions will be handled without a human.

Take banking, for example. Traditionally, a customer would call a line, wait in a queue and be spoken to by a customer service advisor. With the use of chatbots, however, these long queue times can be used for something productive (becoming shorter queue times as a result). Incoming calls can be assessed, allocated to specific departments, and in some cases, even answered through an automated system.

In a recent article from Chatbots Magazine, Sid Garg confirmed that the scope of use for these chatbots needs to be limited. “If your bot is supposed to fully replace a human CSR (customer service representative), it is inevitably going to disappoint customers”

This is, however, not what chatbots are designed for… at least not yet. Instead, they should be considered as a communications platform. Just as email mimicked the postal service (and still does – the email signature is well known as an envelope), chatbots mimic exploratory conversations held by customer service representatives.

The goal, therefore, is efficiency. With more calls routed to the correct place with a quantifiable reason, it’s less wasted time for the people on both sides of the phone. Just like email and post, it takes something that already exists, and has made it go much faster.

Beyond this, however, the end goal is not to replace humans, but to increase the quality of the information we have to work with. In banks, this means fewer calls that are incorrectly routed. In the real world, it’s having technology arrange an appointment, with the only requirement from the end-user being to show up.

Until adoption reaches the point where bots are talking to bots, at least in the consumer sphere, the uncertainty around adoption will still hold sway.

We’re still a long way from the point where AI enabled chatbots can run without any human interaction, but it’s a welcome goal to work towards. The more time that humans can spend doing useful tasks, the more progress can be made on creating and developing systems for the good of all.

Have you ever used a chatbot? Was it up to scratch? Get in touch with us and share your experience.

Read more like this