Get the latest Startup news and information

Please verify before subscribing.

Tech Trends for 2017: Chatbots

The future of customer service is here, and it rests in the artificial hands of surprisingly personable chatbots...

In moments of frustration, jubilation or even confusion, it’s not unusual for people to shout at their phone or computer screen – but what happens when the computer shouts back?

On the to-do list of so many major brands, 2017 will finally see chatbots become the ‘go-to’ for customer service enquiries and online ordering. Described by Microsoft CEO Satya Nadella as “conversation as a platform”, chatbots allow you to by-pass the search box and ask integrated artificial intelligence (AI) your question directly.

While programmes like Siri or Cortana have been around for a number of years, an emphasis on authenticity or human-like attributes is certainly on the increase – and it’s predicted that chatbots will eventually spell the end for apps and manual browsing with your fingertips.

Rather than just an automated robotic voice responding, chatbots will be given personable qualities and even attempt to pass as actual human beings.

Tech business IPsoft has already built a chatbot called Amelia which comes complete with blonde hair, blue eyes and a sharp suit. Described as the “first digital employee”, Amelia is “capable of analysing natural language, understands context, applies logic, learns, resolves problems and even senses emotions.”

Just last June, Amelia got her first job in the public sector after Enfield Council ‘employed’ her to help users complete online applications, including the initial stages of applying for planning permission.

This advancement of AI should come as no surprise to avid tech fans as in June 2014, Eugene Goostman successfully passed the Turing test. The computer programme, which was emulating a 13-year-old Ukrainian boy, successfully convinced 33% of the judges at the Royal Society in London that it was human – though the validity of the result was criticised in some quarters.

This Christmas, Pernod Ricard has launched a chatbot bartender which can help users create cocktails at home by recommending ingredients due to preference of the host, and type and size of the event, and will also direct users to an online supermarket where they can make the purchases.

Just last summer, it was reported, that the “the world’s first robot lawyer” has successfully contested 160,000 parking tickets across London and New York. Created by London-born entrepreneur Joshua Browder, DoNotPay is able to automatically appeal parking fines for users as the process is relatively formulaic – suiting AI capabilities.

How it works

How a chatbot is both created and works depends on how advanced it is. Generally, there are two types of chatbots; one which functions based on a set of rules, and another, more advanced version which uses machine learning.

A chatbot that functions based on rules is incredibly limited and can only respond to very specific commands. You need to be incredibly clear in what you’re saying and use formulaic and relevant English or it won’t know what you mean.

In contrast, a chatbot that functions using machine learning is powered by artificial intelligence (AI). You won’t need to be specific when you are talking to it as it understands language, not just commands, and continuously gets smarter as it learns from conversations it has with people.

Sophie Guibaud, VP of European expansion at Fidor Bank, commented on the future use of chatbots:

“Up to this point artificial intelligence (AI) in banking has been a critical, but mainly ‘backroom’ technology used by financial organisations, especially challenger banks, to unearth deep customer insights and offer them personalised products.

“However, we’re now in an era where customers expect quicker, more responsive services from their banks and this coming year will see the full roll out of AI-driven solutions which aim to meet these heightened expectations, while also ensuring consumers receive the personalised approach they also demand.

“Chatbots, which answer customers’ inquiries and resolve their problems, will become the new norm for banking customers around the world next year. We’ve seen a number of test cases deployed in 2016 but many of the big players including Mastercard will be jumping head first into offering chatbot services for its customers in 2017.

“From an industry perspective, next year will be a big opportunity to really prove how effective consumer-facing AI solutions such as chatbots can be, when applied to everyday customer service.”


(will not be published)