There is a lot of expectation about the potential impact bots will have on our lives. Every major technology company, along with some promising innovators, are fiercely working on their own Artificial Intelligence (AI) offering in the race to become leaders, in what has the potential to become the future of technology. Bots are set to transform the way we use mobile devices, but, more broadly than that, bots could become the most expedient way for people to interact with technology in the future.
Bots offer a contextually aware interface to help users perform functions by understanding the scenario the user is in and predicting what will be useful to the user based on this context. Bots will remove the multistep process a user would otherwise need to complete themselves. The depth and range of information bots will have access to will be exponential and the potential for increased productivity transformative.
Advances in technology, natural language interfaces, natural language processing, neural networks, deep machine learning, along with more powerful processors have helped produce some pretty powerful AI-based voice interfaces. So although the use of voice and user interface is not new, the recent upsurge in offerings has been driven by the advancements in AI-based speech technology. Google, with Google Assistant, Apple with Siri, Microsoft with Cortana and Amazon with Echo are big players in the voice and artificial intelligence space, with their own versions of intelligent personal assistants.
More recently, the announcement of the release of Viv, developed by the team behind Siri, will provide the market with a third-party approach to virtual assistants. Viv will enable developers to build solutions that leverage a conversational user interface (UI) to interact with an app or device. There are also some industrial strength options on the market like Teneo. Teneo, like Viv, can be integrated into third party systems and is already be used by the likes of Shell Oil and Credit Suisse. It has very impressive deep learning capabilities and can understand multiple pieces of information given in a single sentence, it uses context to understand the most appropriate action to take and will ask the user questions to better understand a situation and it can learn about users’ preferences.
Business’ interest in bots and artificial intelligence has also been accelerated recently because of the rise of chatbots and the likes of Facebook, Slack, WeChat and Skype, now enabling businesses to connect with consumers through their messaging apps. Chatbots are using a more primitive version of the AI technology used in virtual assistants and are purported to transform how businesses deliver customer service and engage with customers. As consumers learn to engage with messaging bots, they will expect the same experience from other businesses. We are already seeing businesses starting to build a wide variety of bots to create more personalised and engaging brand experiences. H&M has a chatbot that accesses a users’ photos to learn about their style to then recommend H&M outfits to purchase; and Taco Bell has created Tacobot that allows you to order Taco Bell through the Slack messaging platform, for example.
In the enterprise chatbots can be used to assist users in solving complex business problems and deliver process efficiencies. Currently chatbots are powering human-to-software interaction, but as the technology, and what it can deliver advances, bot-to-bot interaction will be used to deliver even deeper intelligence to users.
The potential that AI and bots present to the enterprise to transform the way we work, the way we engage with brands and most of all, how we engage with technology across platforms is nothing short of revolutionary. However, the AI powering these bots is already available in other implementations and it isn’t working reliably yet. While bots may be able to handle voice recognition and voice synthesis, this is not where the real challenges lie. Understanding the words is not complicated, but deciphering the meaning of a question can be. This will have implications for how bots are used; but whether an organization is building a bot or not, when developers are looking at building any sort of software solution we should all be thinking beyond traditional APIs and thinking about smart bot implementation.
By – Ansible UK