By ROB GEAR, PA DIGITAL expert
Last week I took delivery of an Amazon Echo, which has recently launched in the UK. My initial impression was that although hardly conversational, Alexa the Amazon virtual assistant was quite adept at turning my requests into actions. I was also struck by the ease and enthusiasm with which my seven year old son embraced the technology, bombarding Alexa with questions about football teams and geography.
The ability to interact with machines through direct speech has come on leaps and bounds in recent years – driven by the power of cloud processing and machine learning approaches that have made it much easier for machines to interpret and glean meaning from human speech. Amazon’s Alexa is the latest in a series of virtual assistants including Siri (Apple), Cortana (Microsoft) and Google Assistant (powering Google Home), alongside offerings from other large technology players and multiple start-ups.
In the near term, as with many technologies before, experiences in the consumer space will drive expectations in the workplace. The rapidly evolving and increasingly capable personal assistants underpinned by AI becoming integrated into wearable devices has considerable implications for business.
Just another interaction channel?
For Amazon Echo customers, one of the first things on the agenda is to link the device so it provides a conversational interface to other popular services like arranging a taxi through Uber, ordering a takeaway etc. As adoption of these kinds of services grows, customers will increasingly expect conversational access to be a standard way of interacting with the products and services that touch their daily lives. This will create an imperative for many organisations to ensure products and services can be linked seamlessly to a range of virtual assistants. For many, this may be a far more attractive option than attempting to create their own.
The conversational interface has its own strengths and weaknesses, however. While I was reasonably confident in ordering a taxi with an easily defined journey start and end point, I would be hesitant to book a more complex journey by rail or air, where the range of ticket options and other choices that I would be keen to see would be difficult to present and navigate through speech. For more complex services to work in this format, simplification of service options and/or greater machine intelligence to interpret, adapt and personalise to individual preferences is essential.
Creating new opportunities
As we have learned through Augmented Reality (AR) products such as HoloLens and Google Glass, the combination of direct interaction with human senses through untethered, hands-free and wearable devices is a powerful combination.
In this case, a powerful virtual assistant or AI-enhanced expert system that can be interacted with conversationally suggests a host of exciting new applications in areas such education and learning, or any situation in which a worker may wish to describe a problem and then receive help and guidance, while keeping their hands free. Future combinations of visual (via AR) and audio (via a conversational AI) support may transform the way that all kinds of complex tasks and experiential learning are carried out. Combine all of this with more intuitive gesture controls (for example Bragi’s ‘The Dash’), you have a powerful new model of human computer interaction and one that is geared much more towards future models of symbiotic working between people and machines.
An adaptive technology
The cloud-based machine learning conversation engines that underpin the new generation of virtual assistants are highly adaptive, and over time, become increasingly personalised to the nuances and idiosyncrasies of an individual's speech patterns and conversational style. They provide an opportunity for organisations to deliver far more fluid and intimate personal experiences as well as new opportunities for learning more about customer preferences. What is less clear is who owns the data resulting from the conversational systems learning an individual's preferences? What would happen if I wished to move from Echo to Google Home? Would it be possible to migrate the ‘knowledge of me’ between services? Could I request all customisation data to be deleted?
A new frontier for privacy concerns
The conversational interface is not without its risks. At face value, the very idea of having a live, networked, corporate-controlled microphone in the home will immediately trigger fears of potential invasion of privacy for some, and consumers will have little tolerance for any security breaches or misuse of data in relation to these types of service. The more intimate and personal the model of engagement, the greater the expectations of rock solid security and ethical behaviour will be, and the greater the damage to brand and reputation should any mishaps occur.
So what’s next for conversational interfaces? Developments continue at pace and we are already seeing the arrival of exciting next generation capabilities. Viv deploys AI techniques and the ability to generate new code on the fly to answer complex conversational enquiries such as “will it be warmer than 70 degrees near the Golden Gate Bridge after 5pm the day after tomorrow?”
What might take a little longer is the social change towards people feeling confident in engaging conversationally with machines. If the experience continues to improve and the convenience and value of speech interfaces are rapidly evident, we could see many more people forsaking their keyboards and touchscreens in favour of dialogue with personal assistants, concierges and voice assistants just like in the movie Her. The big tech companies all appear to betting quite heavily on this outcome.
We’ve recently installed an Amazon Echo in PA’s iLab and will update you with our impressions as we explore under the bonnet of the technology. In the meantime, how do you feel about speaking to AI bots and what kind of uses for conversational interfaces can you see in your organisation?