Chatbots and Voice Assistants are moving towards each other
‘Hi Zeph.’
‘Welcome to your car, Penelope, where can I drive you today?’
‘The nearest café please.’
‘Certainly.’
‘And Zeph?’
‘Yes Penelope?’
‘Please lower the air con one step and tune the radio to a rock channel.’
‘With pleasure.’
Imagine a hands free driving experience, where your interface with your vehicle is entirely conversational. You can relax, sleep, be entertained, or do some work. Our time is increasingly under pressure, so it would be wonderful to gain some back in this way. And this scenario is no longer one in the realm of SciFi. Most of the technologies that are required for it to be realised are either in place or in advanced stages of testing.
The one I want to focus on here is that of the conversational interface. Thanks to Siri, Alexa and Cortana we are becoming used to the practice of using conversation to direct technology. And there are tremendous advantages to being able to do so. Obviously, having your hands free and your vision liberated from a small rectangular area is a huge physical plus.
Another gain for humanity from this technology is that it plays to our strengths. Humans are good at talking. It’s one of our defining characteristics. Whereas learning how to control new technologies that require button-pressing or symbol-using navigation is not intuitive. Anyone, young and old, can carry out a conversation, while not everyone has the time or energy to learn a new systems: or even the complexities of a modern TV remote control!
When humans can control the devices around them — especially computers — through conversation, we will be living in a world where seeing someone confined to a desk or using a keyboard and mouse will look as dated as watching films where characters have to run to a booth to make a phone call.
Some technology experts have been at pains to emphasise the difference between speech-to-text technology and chatbots. This seems to me to be placing a false barrier between the two and it is surely in the synthesis of both — already closely-related — technologies that this future lies. I spoke to Avi Ben Ezra, CTO of the award-winning SnatchBot company, which has one of the best AI chatbot platforms, to dig deeper into this issue.
‘When we founded SnatchBot,’ explained Ben Ezra, ‘we made a strategic decision to think long-term. Where will chatbots be ten, twenty years from now? Our answer: everywhere. This is provided that chatbots are not confined to any particular channel (like Messenger, or websites, etc.) and provided they are adaptable, so that they suit every vertical.
‘It was clear to us at the outset that text-to-speech and speech-to-text was going to be a crucial development in the evolution of chatbots, that people needing to tap at keypads or peer at screens was only a passing phase of interface technology. So we have a roadmap that leads towards the complete chatbot.
Build better voice apps. Get more articles & interviews from voice technology experts at voicetechpodcast.com
‘Speech-to-text technology is coming on fast and text can be fashioned out of the verbal statements of humans with near-human accuracy. To fulfil its potential, however, this technology needs to be smart, it has to integrate with advances in AI and NLP, so that not only can software transcribe what has been spoken, it can understand speech.
‘Or to look at it from the other way around. Now chatbots can identify entities with nearly one hundred percent accuracy and intents with constantly increasing effectiveness, they need to be able to listen and speak to be the tool that releases humans from equipment that is going to look very primitive to the next generation, whether it is a steering wheel or a mouse.’
In September 2018, SnatchBot announced a significant moment had arrived on its roadmap towards this hands-free future, with the provision of a text-to-speech functionality, made available for all its chatbots, whether the free version or those built on the company’s PRO-plan. Coincidently, that month too saw the chatbot platform add Viber to the long list of channels on which SnatchBot’s bots can be launched. Impressively, you can start chatting to the bot on a website, swap the conversation seamlessly to Messenger, move on to Skype, Slack, LINE, Telegram etc. and now Viber, a channel with 900 million active users.
The fact that one and the same chatbot can be channel agnostic as well as smart and capable of listening and speaking, is the reason Ben Ezra is confident that chatbots are going to be central to the revolution that is taking place in the application of AI. ‘The conversational interface is already present in a myriad of use cases. There is a tipping point approaching where the public are going to find intelligent, conversational chatbots everywhere and this will be a cause for delight. People who are intimidated or averse to technology will discover nothing more difficult than an ability to talk is needed to have full control over the technology around them.’
