Artificial empathy - the next step in mental health tech?

By Martin Vowles - 08 Jun 2017

For the past 5 months, Plexus and Money and Mental Health have been working together to beta test The Shopper Stopper, a browser plug-in to make your self-control stronger and help you to think before you click.

Impulse buying is an emotional response. We rarely buy something we genuinely need with little thought, online, late at night. We buy things we think we want, perhaps to ease a little boredom, loneliness, stress. The Shopper Stopper allows users to identify the times of day when they’re at risk of impulsive spending and ‘close’ their access to online shops, restricting when they can make purchases through selected sites.

It’s a great resource to help people start managing their spending and during the test we’ve seen some amazing results, reaffirming our belief that there can’t be a one-size-fits-all solution to any health issue, especially in mental health.

Everyday AI

Our daily lives include constant interaction with technology – Alexa, what’s the weather going to be like today? Siri, play the latest Radiohead album. What’s for dinner tonight – shall we order a pizza online? .

Massive corporations like Amazon, Google and Apple have brought artificial intelligence and machine learning into our homes in a way that’s made it second nature to technically engaged users. It’s become intuitive, in the same way that my 3 year old will touch any screen and expect it to interact with them.

Every day, mundane tasks like ordering food and shopping for household necessities is done through technology and even appealing parking tickets can be managed for us by chat bots like DoNotPay – the ‘world’s first robot lawyer’. The creators of this technology have made things easier and easier for us as they seek to reduce ‘friction’ – one-click shopping, mobile banking, integration with other apps like maps or creating tickets on your phone. Everything is getting quicker and easier.

But while there’s no doubt that reducing friction creates a better customer experience, is it good for us? It was much harder to give in to impulse shopping while bored or lonely when all your local stores were shut at 6pm. Seamless experiences can leave us open to indulging in damaging behaviour as we ease our emotional state by triggering the quick-hit, endorphin release of online shopping.

If only I had a heart...

We’ve traditionally held a sort of master-slave relationship with our tech. We ask it to perform a task for us and it completes it. Is it now time to deepen this relationship? Can we invest our trust in artificial intelligence to help manage our behaviour? A kind of artificial empathy?

Plexus have also been developing a range of tools to help people nurture their mental fitness, through a developed process of ‘reflect, review, learn’. Grace takes five minutes a day and can help users identify patterns and stress points in their working week; Tuesdays for example, might be particularly stressful due to a change in commute or workload in the office. They may also discover that Tuesday evenings are when the bulk of their impulse purchasing is done and can connect the two. In this way, we hope that people are able to see a clearer view of their emotional health and gain help and support to take steps to address these challenges. Our little digital cup of tea.

Could technology like Alexa and Google Home do more to combine these approaches? It should be relatively simple for Amazon to introduce an optional function which limits users’ ability to complete purchases at certain times of day – redirecting items onto a wishlist or to the ‘save for later’ section in the shopping basket, and disabling one-click purchasing. Could it go further? Could machine learning allow it to identify purchasing patterns and to recognise the items as ‘want’ rather than ‘need’ purchases? Perhaps even suggesting other ways for the user to distract themselves, by watching a new series on Amazon TV or listening to a recently streamed album?

And do we trust them with our data?

The key to this kind of interaction is trust. We, as users, would need to develop trust in the companies we buy from to honestly help us manage our behaviour and not see them as solely interested in parting us from our cash and stealing us away from competitors. A big ask. We would also need to trust that our information and our privacy would be respected. Introducing Data Licences – something all users of Grace set up – giving them total control over their data and how it is used and shared with Plexus. It is renewed periodically so that users know what happening with their interactions. It’s a great first step to build this kind of trust and allow systems to be built up around them.

Naturally, The Shopper Stopper needs further user testing and refinement. Future iterations could utilise A.I. so that it can become a tool which helps to identify your positive and negative online behaviours, for example, via access to a user’s browser history or cookies being added by sites visited. This extra data should allow The Shopper Stopper to become a supporter who can give a gentle nudge when something may be amiss or when it sees trends that could be harmful to you in the long term.

We don’t have all the answers yet but we are just at the start of a journey to discover how we can evolve and manage our emotional relationship with artificial intelligence, and the new technological opportunities it will present in the coming years, not just to make life easier, but to make it better too.

Our trial of the The Shopper Stopper is coming to an end, though we’re keen to talk to organisations who’d be willing to take it on as a long term service for those who need it. Learn more about The Shopper Stopper.