Designing personalities — event poster made by Designit

A summary of thoughts inspired by Ladies that UX Aarhus meet-up at Designit. With an agenda that was as immersive, as it was engaging, we had a great afternoon of theories, case samples and discussions, that made us more aware of designing AI personalities responsibly.

The emergence of Artificial Intelligence experiences creates possibilities for new ways of interaction between user and product. This in turn, puts new demands on us as user experience designers. Because, how do we design intuitive voice or visual interfaces that give human-like experiences? And how do we ensure that we do not just create a copy of Siri or Alexa, but that it becomes a true extension of our brand? These were some of the questions discussed at the event.

We have all been hooked by the buzz of artificial intelligence, be it from our favourite science fiction movies, Apple’s Siri, Amazon’s Alexa that are at our service when we ask them to, or the new Google Assistant presented in Google I/O 2018 conference, that will be able to act on our behalf.

It is exciting to think about the ways this technology can make our lives easier, but equally frightening to experience the pace of it’s development.

The question most people ask is, what’s next? Will robots take my job, or raise my kids? Not quite yet, an article from the Guardian writes. AI is supposed to free us from mundane jobs, rather than become our competitor.

Ok, so now that we have established that AI is here to stay and feel somehow safe from the threat of robots taking over, how do we design them to be as human as possible? Should we humanise them, or should it be obvious that we are communicating with a robot?

Studies of human-robot interaction show that we favour human-like appearance and behaviour of chatbots/robots, rather than mechanical ones. Still, the designed human trades designers give to robots, and what an average user perceives as being human, can be very different.

“When we design a new AI we need to be very selective about what traits we choose to include or exclude in the personality we are designing for AI.”

In her presentation, Stine stressed that as humans, we will always project our emotions to the objects around us. For example, if you have one of those robot floor cleaners, you feel some kind of affection to it, because it cleans your house and that makes you happy. But, what exactly do we need to be selective with, when we design these personalities?

Frederikke’s take on it, was to have a look at human personality studies, and draw theories we know from there, to help us design personalities of robots. For instance, deciding on the level of extraversion of voice interfaces can be tricky, when such interface might be designed both to interact with children and sick family members.

There was a very nice energy when the audience got the challenge of designing the personality of a chatbot for a business, which can commonly be perceived as conservative. They jumped right at it, and shared their ideas.

Several things were taken in consideration, such as who will be the user communicating with this chatbot? What are some tasks the chatbot will help solve and should the bot be friendly or formal, male or female?

Apparently, most of the groups agreed on a male in his forties, which can be understood considering the brand it was being designed for. And since there were many discussions about personalising the communication experience, it made me wonder, how much of our data should these companies and designers use to create a personalised experience, and where do we set the privacy boundaries?

I am very concerned about how much of my personal data companies get, save and use on my behalf, in the name of giving me a more personalised experience with their services. Not only because it is my right to choose what I share with the public, but because I know they can use this data to manipulate my decisions — sometimes by using friendly bots as a medium.

I think as people, and especially as designers, we should be more aware of how the hard work we put to design our services is used. Be very concerned about dark patterns — for instance letting a bot make decisions in the user’s behalf based on an algorithm of behavioural tracking. How will this be used by corporations, who want to sell more products? We have to choose to make a stand for the user’s (people’s) privacy consent as soon in the design process as we can.

All in all, LTUX meet-up at Designit was one of the best events we have hosted so far. With a rockstar presentation from Stine Skaarup and Frederikke Tholander from Designit, fun exercises and engaging discussions, it couldn’t have been better. We did not find the answer as to how to design a perfect robot personality, but we now know the right questions to ask when put in front of this challenge.

Our monthly meet-ups with Ladies that UX Aarhus are a source of knowledge, inspiration, good energy and networking. If you want to know what Ladies that UX Aarhus is, read the medium post I wrote about creating this community. If you are located in Aarhus, you can also join our meet-up group to experience it directly.

Full-time human, with a pencil and something to say.