It’s a regular occurrence in our household to see those in their 80s sat next to those in their teens, as they both simultaneously speak into their smartphone to search Google via Alexa, Siri or similar.
It’s a perfect example of how, regardless of generation, we are all adapting to the ‘robot’ in our palms.
Machine Learning, Deep Learning and the other more specific applications of Artificial Intelligence (AI), such as Natural Language Processing, are becoming commonplace. Whilst we may not be able, as customers, to pinpoint the tech that’s being used, we are encountering more and more applications in our daily lives.
As a career technologist, with a breadth of experience across a number of areas, I see trends growing out of one sector and across into others.
Insurance is no different - from the underlying risk models where vast quantities of historic data can be used to elicit patterns of risk that no single person could identify, through to the use of consumer behavioural data which can help predict buying habits.
The future is Machine Learning
AI, and specifically Machine Learning applied to structured and unstructured data, is being used to create commercial edge in existing businesses and, in some cases, innovative products or business not previously available in the consumer market.
Global giants have the investment breadth to seek out dominance in the use and creative adaption of AI techniques, however it’s partnering with sector specialists that will enable the best marriage between pure tech AI and commercial or consumer applications.
Our business is the distribution and servicing of financial products, with car insurance at its heart. Car insurance is something which most people can relate to. We have all experienced the changes in how we drive, how we insure and the in-car safety systems we all have come to rely on.
B-call and E-call buttons in our cars are becoming more common, delivering rescue and recover co-ordinates to the services when we most need them. Combining satellite data, telemetry data, and sensor inputs from the car to diagnose the situation. Voice-triggered and prioritised based on the data being transmitted, we are getting used to being connected directly with breakdown services.
Cars will already automatically brake for you, they will sense if your attention wanders and take action, they will even park themselves for you. The car will run diagnostics, wipe the windscreen and put lights on when needed. The level of automation (robotics) in our cars is already extensive, and it’s growing.
Telematics products - all the rage a couple of years ago - have largely failed to take a hold in the market. A combination of the awkward installation process when you buy a telematics product perhaps, or the calibration of the user experience as it appears in the eyes of the young driver when they feel penalised for one subpar drive, or even the general driving public wondering why the car in front is adhering religiously to the speed limit rather than stretching the boundaries.
The point is, that the more data available the more likely a business is to really understand it’s customers behaviour, the cause and effect of actions and the genuine product experience a consumer has, rather than the white boarded experience it thought they should have.
The power of the voice
Customers speak a lot. To each other, over the phone and now more often to devices in their homes, be it smart assistants, TVs or other voice-activated applications. This is possible as the technology which underpins voice recognition is rapidly developing with access to the large-scale data and computing power required.
The current generations of Natural Language Processing, sentiment analysis and, to an extent, voice biometrics, enables simple targeted conversations to be designed to help consumers get to where they want to be quickly and naturally. Dovetail a human experience into the mix, you’ll not only get to the services you want, but the conversational ‘bot’ gets a little better trained at every interaction.
Speech Recognition modules available from the big tech players can struggle with some of the unique aspects of the British systems, but they can learn. Using the data at our fingertips, we can adapt them for our country, our industry, our products and our business. Customer conversations, intelligent and responsive answers, tailored for you is within our reach. I don’t mean that every conversation should be automated - absolutely not we must offer our customers the best of both worlds; the immediate and always-on service offered by a robot for the simpler, more predictable conversations, combined with humans to respond to the more complex queries.
How is this changing the nature of our business? The data and algorithms we use enables us to prevent fraud. It helps our insurer partners minimise losses and it ensures that the product we offer is affordable. We get to aggregate competitive prices from a wide range of suppliers, ensuring that customers have access to the most keenly priced product for their needs. Overall, this means less waste in the insurance market and better value for customers.
Applying these same customer centric principles to AI, via our voice assistance and conversational design, means that it should be as easy in the future to insure your robot, as it would be to insure your car (which is already part robot).