Artificial intelligence (AI) has been widely considered as the buzzword of the tech industry in 2017. The UK government has even dedicated £75m of funding for AI, including up to £45m to build AI capability and knowledge by increasing the number of AI PhD students to 200 a year.
With governments launching dedicated funding and organisations from all walks of life racing to reposition themselves as tech companies - from launching chatbots to deploying big data analysis and deep learning - AI is transforming the way we work and how we interact with brands.
In today’s economic climate, organisations are seeking alterative methods and technologies to serve a larger volume of people, keeping customer service standards up and ensuring costs are managed. It is in this area that the explosion of AI-powered chatbots has disrupted the very meaning of customer service. Nuance’s recent research found 89% of consumers admitting they prefer to engage in conversation with virtual assistants to find information, rather than searching through web pages or a mobile app on their own.
Another recent study by Twilio, echoed these findings. Surveying 6,000 consumers in Europe, Asia and North America, the poll found that nine out of ten consumers prefer to use messaging to talk to businesses than all other channels.
User experience and ease of use are, of course, among the top priorities for any product. And the rise of the conversational digital assistants, like Google Home and Amazon Alexa, has demonstrated the power of deep learning and speech recognition for creating devices that provide an intuitive way of engaging with online services and content.
Amazon and Microsoft chatbots combine
With advanced chatbot technology fast influencing the smart home, it’s not just companies driving their own agendas forward today. Just recently, Amazon and Microsoft announced their collaboration - getting their virtual assistants to talk to each other. This partnership will mean Alexa and Cortana - the two voice-activated AI assistants from the respective companies - will be able to communicate with one another to access their individual features. For example, someone with an Amazon Echo smart speaker might use Cortana to book a meeting or to check their calendar.
Cortana isn’t the only VA to partner with Alexa. In fact, Nuance has recently launched technology that does just this as well. Its Nina for Amazon Alexa VA is the first intelligent enterprise virtual assistant that integrates with the popular internet of things (IoT) device. Nuance delivers Nina as an Alexa Skill, allowing organisations to leverage their investment in Nina to engage through Alexa-powered devices.
This brings to the fore an exciting question: how exactly do two virtual assistants, two AI solutions, work together? The AI solution must know what it does not know, and what the other does know. Alexa should know what Cortana knows to be able to decide who ought to respond to a specific inquiry - and vice versa.
Of course, in a simple scenario, the user can explicitly instruct a virtual assistant to consult with the other, such as "Alexa, ask Cortana if I have any appointment today". However, it would be much better if both interoperable AI solutions knew by themselves who can solve the problem. Nuance's Intelligent Knowledge Access technology can already do this for passive back-end knowledge sources. It knows what data and knowledge databases it needs to consult to answer complex user queries.
In the future, this could also be extended to include meta-knowledge about other AI solutions and assistants. In the best case, if this knowledge is not static, the assistant learns from experiences and develops further. Depending on how well the interactions with other virtual assistants work - and how satisfied the human user is with the results - the model can be refined over time.
It’s not just virtual assistants talking to each other that’s important, of course. In the future, successful digital assistants will need to be capable of holding an intelligent, two-way conversation with the user. Like a human, it will have to maintain context as the user changes subjects or uses colloquial, conversational expressions and words.
Today, most digital assistants aren’t sophisticated enough to do this. While some can successfully respond to a single, basic inquiry, such as, “what is the temperature in London?” (answer: “It’s 15 degrees in London.”), if the consumer follows up with the conversational, “how about Bristol?” most digital assistants today will not understand the context that the question is about the weather.
Harnessing this knowledge of context is a clear challenge and focus are for conversational AI leaders, as humans do not communicate in direct, full sentences when naturally conversing.
VAs needs to know who’s talking (and who to listen to)
When a recent Google ad for Google Home made thousands of virtual assistants react to what they thought was a command - and when the same happened to Amazon Alexa ordering doll houses - it made one thing clear: even when assistants have a wake-up word or phrase, this is not sufficient to ensure they only react to their owner. But they absolutely should have the capabilities to do so.
Technology can help here: Enhanced with voice biometrics can ensure the device only listens to its ‘master’s voice’ for example. Consider how powerful this could be in a family setup, where a device can make sure not everybody has permissions to purchase things, for example. This is in addition to other use cases of biometrics where it is used to verify a person is who they claim they are, e.g. when calling a bank, instead of entering PINs or answering security questions. Identifying the speaker also enables personalization opportunities for brands, identified through learning about an individual’s preferences.
An AI-enabled future
From our homes to our workplaces, and everywhere in between, AI is changing the very face of how we communicate, work and consume products and services.
The technology is disrupting virtually all industries, from healthcare, finance and customer service to automotive and telecommunications. The sheer power and intuition of the technology is enabling organisations to build closer relationships with their customers, through improved learning to natural, super-powered customer service, armed with knowledge and insights. With this new, intelligent technology now communicating with each other as well, we could be on the cusp of another great new beginning.
Nils Lenke, Senior Director, Corporate Research at Nuance Communications
Image Credit: Montri Nipitvittaya / Shutterstock