Should We Give Chatbots Their Own Personalities?

by Nikita Johnson

Today, we have machines that assemble cars, make candy bars, defuse bombs, and a myriad of other things. They can dispense our drinks, facilitate our bank deposits, and find the movies we want to watch with a touch of the screen. Automation allows all kinds of amazing things, but it is all done with virtually no personality. Building a chatbot with the ability to be conversational with emotion is crucial to getting people to gain trust in the technology. And now there are plenty of tools and resources available to rapidly create and launch chatbots with the personality customers want and businesses needs.   Jordi Torras is CEO and Founder of Inbenta, a company that specializes in NLP, semantic search and chatbots to improve customer experience. We spoke to him ahead of his presentation at the Virtual Assistant Summit in San Francisco, to learn about the recent explosion of chatbots and virtual assistants, and what we can expect to see in the future. This year has been called "Year of the Bots" - what do you feel are the leading factors enabling recent advancements and uptake of chatbots and virtual assistants?Neural Networks were invented during mid 1980's, but they fell into oblivion due to some disappointing results when trying to solve problems like computer vision. However, after 2012, deep feedforward neural networks developed in the research group of Jürgen Schmidhuber at the Swiss AI Lab IDSIA triggered a whole new revival of machine learning applications. Basically these new neural networks work based on different, subsequent layers of abstraction, each one specialized in a particular task, and that has dramatically improved results in several applications. That has produced resurgence of all fields traditionally related to AI, like computer vision, gaming, natural language processing, and therefore chatbots and conversational-based applications.   What present or potential future applications of virtual assistants excite you most? Nowadays we interact with government, corporations and services basically by 3 methods:

  1. Physically interacting with people (ordering food in a restaurant, purchasing in physical store, talking to a bank teller in a physical branch etc)
  2. Talking to people over the phone (basically calling a call center)
  3. Typing (and clicking) on a computer (websites, enterprise apps)

Because computers have been historically really bad at understanding human language, we had to get used to computer keyboards, mouse, screens, touch screens and a variety of ways to get the job done, all of these methods are new to humans, and therefore difficult to learn. What most excites me is that humans will be interacting with companies and services through conversations (like method 1) but the whole job will be done by computers (like method 2), completely reshaping the way we interact.   In the future, which industries do you feel will be most disrupted by virtual assistants, and AI in general?Mostly big B2C companies that have big call centers and also big traffic on their websites or mobile apps will be the first ones to move to AI, chatbots and virtual assistants. That includes a wide range of vertical sectors (telco, e-commerce, banking, insurances, and many more).   In the future all businesses and governments will have their interactions through some sort of AI-based dialog system, as these systems become easier to use and more affordable.   What developments can we expect to see in virtual assistants in the next 5 years?I believe that chatbots, voice-recognition systems and generic assistants like Siri or Cortana will eventually merge into one cohesive user experience. We don't have yet the "google for bots" but eventually we will, and we ask Siri, she will be able to forward the question to the appropriate bot, and answer my question effectively. We are now halfway, where Siri basically uses Google (or Bing) to answer all the questions that she can't answer directly.   What is your main goal for Inbenta?I want Inbenta to be the company that brings conversational intelligence to the enterprise, that completely reshapes the way we interface with organizations, in a way that we remove forever the pain of interacting with ill-designed websites and overflowed call centers. So, like many other tech startups, we want to make the world a better place! Jordi Torras will be speaking at the Virtual Assistant Summit in San Francisco on 26-27 January! Other speakers include Roberto Pieraccini, Director of Advanced Conversational Technologies at Jibo; Anjuli Kannan, Software Engineer at Google; Alonso Martinez, Technical Director, at Pixar; Milie Taing, Founder of Lili.ai; Matt Schlict, Editor of Chatbots Magazine; and Lionel Cordesses, Innovation Project Manager at Renault.Discounted tickets are available until 7 October, book your pass at an Early Bird rate and view more speakers on the event website here.