Towards Turing: personality without people
The Future Foundation’s Richard Nicholls takes a look at the part that Artificial Intelligence plays between consumers and brands, and explains why today’s AI so often does little to convey the personality of the brand behind them.
“I believe that in about fifty years’ time it will be possible to programme computers… to make them play the imitation game so well that an average interrogator will not have more than 70 percent chance of making the right identification [between computer and human] after five minutes of questioning.” – Alan Turing, 1950
Artificial Intelligence has already become a part of so many everyday relationships between consumers and brands. Many supermarkets, airlines and banks, to name just a few examples, already make use of automated chatbots, voice-recognition software or self-checkout technology. We talk to the machine and the machine talks back – but often, not very well.
Today’s AI so often does little to convey the personality of the brand behind them, limited to specific and obviously functional contexts. It is still far away from truly convincing conversations or linguistic intelligence (no matter how sophisticated their logic).
They can make themselves useful, and the data-crunching capability is especially powerful with more behavioural data available and consumers demanding ever better results, more information, clever analysis and creative, personalised recommendations.
But they are far short of the Turing Test – computer science pioneer Alan Turing’s ‘Imitation Game’ where a human asking questions finds it difficult to distinguish between responses given by a machine and those by a human.
We hate being treated like parts on a production line, or lines in an algorithm – and get frustrated when we are inconvenienced by an AI that lacks the emotional intelligence or common sense to work through a simple problem. But we also demand speed and efficiency in customer service.
There is a conflict here: we want speed and humanity. If these traits are currently seen as mutually exclusive (a feeling that computers lack the human touch), the trade-off is becoming smaller. Between 2009 and 2012, the number of British consumers who saw self-service’s inanimate nature as a benefit nearly doubled, to 32%. Can self-service’s strengths – its speed and consistency – end up making it more attentive, even more empathetic, than the humans it replaces?
“What do you think are the general benefits of self-service?”
Source: nVision Research/Monitise | Base: 1,000-5,000 online respondents aged 16+, GB
We expect future systems to operate as true brand ambassadors – true Artificial Personalities. The point isn’t really to pass the Turing Test but to come close enough not to make the human-AI interaction grating; to have enough linguistic intelligence to complement the logical intelligence; and to be equipped with enough of the individual’s behavioural data to build a proxy for emotional intelligence.
Brands that have cultivated a unique identity will have to learn how to project it through the proliferating interfaces of our digital world with an Artificial Personality of their own.
Of course, the last few years have seen huge advances in digital assistant technology. Between 2008 and 2013, intelligent digital assistants such as Siri and Google Now gained over 400 million users. The question for brands is how to push their own brand identities through digital personalities inhabiting future smart devices (and wearable technologies).
Automated customer service portals can be little more than well-dressed encyclopaedia at present. In future, a sense of artificial empathy will differentiate the true brand avatars building marketing interactions and making an emotional connection with customers. We expect to see the balance between human and automated customer service personnel shift as technology evolves: as Artificial Personalities become more human.
The in-store environment could make much of this trend. And there are plenty of interesting examples from around the world that we have noticed at Future Foundation.
For example, Toshiba’s next generation of self-checkout terminals, set for 2014 introduction, can visually identify fruit and learn from any mistakes it makes. How long before self-checkout gets even more intelligent, initiating discussions and answering questions?
Japanese retailer United Arrows used an Xbox Kinect to create two mannequins that mimic the actions of passersby in the shop window. The technology affords customers a unique view of clothes on a moving body. (Retailers would here need to be mindful of the ‘uncanny valley’ effect – when mannequins or robots look and act almost, but not quite, like humans, it can be very disquieting.)
In the future, we expect The Artificial Personality to become far less contextual. Today’s avatars might produce convincing acts of insight in a highly specialised field. Future artificial personalities will be equipped with a greater understanding of language and will be able to make inter-disciplinary judgements. (Using medical expertise, for example, to make recommendations on adventure holiday plans.)
And, with wearable technology presenting new digital routes to consumers, the imperative to project an Artificial Personality grows stronger.
Where else do we think the Artificial Personality trend will go next? We expect more sophisticated Artificial Personalities to become more important in communications and customer service as greater emotional intelligence complements the traditional benefits (e.g. speed, data) of artificial intelligence.
AI in the in-store environment will answer more complicated questions and give more personalised responses. Maximising services – those giving recommendations – will become better able to offer tailored advice for more complicated parts of our lives – our leisure time or our media consumption, say.
Treated as an opportunity to express brand identity, rather than just an automated access point, Artificial Personality avatars will become powerful brand assets.