Return to site

Three challenges we must overcome to unleash the potential of AI

By Kent Mackenzie, director, Deloitte.

When the phrase ‘artificial intelligence’ (AI) is mentioned, many people have visions of cyborgs, automations, or even a replicant from Blade Runner – the latter if you’re of a certain era. It’s fair enough: many films, documentaries and magazines have conveyed the concept of artificial intelligence in a robotic form. 

While that sci-fi fantasy hasn’t yet come to pass, it might well be the case in ten to fifteen years’ time. But no matter how it looks, or the exact form it takes, AI is becoming an increasingly hot topic of conversation in a variety of industries and circles.

Many argue, particularly in the financial services industry, that AI will never have a physical manifestation - there won’t ever be a time when a robot at the end of the phone helps you with your pension queries. Instead, it will be machine-based deep learning and cognitive computing, pulling in a staggering amount of data to help those who use it make better decisions.

To the man on the street, it can often be hard to imagine how this works in practice. But similar concepts are already being used by your average person every day. Online robo-advisors and chatbots, which can offer customers simplified advice or guidance about a variety of products and services, are using the same kind of processes and technology to deliver real-time information.

It’s the next industrial revolution happening right in front of us. AI is going to open up more options to customers and businesses that the human brain would ever be able to analyse on its own. The technology allowing this to be put in place on a wide scale is being developed all the time, but there are still challenges to overcome.

The first is humans’ readiness for this kind of change. Taking the financial services example, many people still feel uncomfortable making important decisions about their pension based on what a machine is telling them. There’s a battle to win hearts and minds, particularly with generations used to working in traditional ways.

To overcome this, it’ll come down to education and relying on the next generation to help it grow. Time is increasingly precious to people and the immediacy and benefits of services using AI will become hard to resist. There will be a move to bionic advice, where human interaction is complemented by AI automated services, to improve the quality of guidance, and lower the cost to the customer.

In FS, it can allow for people to have a “financial butler”, automated to help manage all aspects of their financial life. For example, the iOS and Android app, Chip, uses an algorithm to calculate what its customers can afford to save based on their spending. It then transfers the money from a current account to the Chip savings account automatically.  Imagine a world where your online AI-based ‘butler’ is continually switching the product or service you consume to make sure you are receiving the best deal or service – the reality of this could soon be upon us.

The second stumbling block in AI’s development is a lack of readiness from businesses to weave it into their processes. In many cases, their data is stored in a variety of different places and operational structures that make it challenging to integrate AI quickly, and without significant cost. The onset of cloud-based infrastructures and smarter ways of scraping data will help overcome this, but no doubt there are still a few legacy issues to untangle before advanced analytics and AI can become commonplace.

Take Insurance companies for example, they are likely to leverage AI far more in the future to better understand the behaviour of their customers and, in turn, create things like ‘real-time’ car insurance premiums based on where and when they’re driving, what the weather is like that day, and a variety of other variables. The onset of the connected world, and the ‘internet of things’ will also likely be integrated into AI to revolutionise the entire general insurance market.  Whilst we begin to see this come through in innovations such as Aviva’s driver application, they are still early in their evolution with great opportunities on the horizon.

That’s a lot of data to keep track of, which throws up potentially frightening questions over what is monitored, how, and by whom. It’s no surprise then that regulation and ethics is the third issue we need to consider carefully as the adoption of AI and democratisation of data takes off.

Data and information can’t flow freely without some form of control. There’s a need for it to be used responsibly, ensuring businesses cannot monitor every move their customers make and information is held securely.

Data changes everything – so could AI. But we need to know that we’re introducing it for the right reasons and to deliver the right outcomes. Nevertheless, an AI revolution is on the horizon and it will have consequences for everyone: on our everyday lives and the way businesses work. It’s time we began to embrace the change. 

Deloitte is a silver sponsor for Data Summit 17. To find out more about Deloitte, join their talk and visit their stand at the event.

All Posts

Almost done…

We just sent you an email. Please click the link in the email to confirm your subscription!

OKSubscriptions powered by Strikingly