As a submission to the 2019 Microsoft Design Expo, we were challenged to design for “Empathy at Scale”. Our team wanted to tackle the problem of creating empathy by targeting the obstacles we face in becoming open to different perspectives: our emotions and biases.
Advanced IxD Course
University of Washington
We believe empathy is built through considering others’ perspectives and emotions, often through imagination if we haven’t had a similar experience.
However, to even start the process of empathizing, we first need to process our thoughts and emotions in order to look beyond our biases. This is what we call Reflection. Without reflection, these emotions become obstacles to empathy.
By defining what empathy meant to us, we began by outlining how we are able to open up to understanding different, maybe even opposing, perspectives. We wanted to start from when we encounter internal or external conflict, and the process of reflection.
We discussed a basic framework of reflection:
1. Venting: the opportunity to express your negative emotions
2. Recounting: examining the situation retrospecively & carefully
3. Reflecting: realizing our own biases & considering other perspectives
This process isn’t something new—sometimes we’re able to do it successfully ourselves (or with our moms who are always right). We believed that by creating a safe space to practice reflection could build a habit of empathy, using our own memories and thoughts as a mechanism for remembering.
We wanted to be able to address conflict from both sides whether it’s someone who may not have the time and space to process their own emotions or someone who may need that extra bit of guidance to realize their own biases.
After identifying our target users, we storyboarded how two users involved in the same conflict would use Ponder and come to an understanding of eachother. Creating a story around the user flow helped us identify the main interaction points of Ponder that we would need to explicitly design for.
To quickly test our concept, we conducted a wizard of oz session with three participants. Through a complicated mastery of multiple computers and a blutooth speaker hidden in a tissue box, we asked participants to go through the activity of reflection by theirselves, with a friend, and finally Ponder (aka us).
Important takeaways from the research included:
1. Onboarding: users needed some time to get to know Ponder to feel comfortable confiding in it.
2. Feedback: similar to other conversational UI devices, Ponder needed to give visual or auditory feedback to the user that it was active.
3. Escape route: it was important to recognize that sometimes conflict resolution and reflection wasn’t something that could happen over one day. Users needed an escape route when they felt too overwhelmed or tired of the subject and Ponder needed to recognize when to stop.
For our final presentation, we wanted our video to spend less time explaining Ponder and more time showing the different perspectices, stories and emotions that we can all feel dealing with conflict. Though I wasn’t directly involved in the filming, the rest of us contributed through planning, casting, and carrying a lot of camera gear (me).
A conversational AI that helps individuals realize their own biases, process their own emotions, and open up to other perspectives.
Without a presence like Ponder to help guide reflection, people are often unable to process their emotions in a productive manner that hinder our ability to be open. Additionally, sometimes those close to us aren’t the best suited to help guide us through productive reflection. In order to avoid conflict themselves, they often resort to validating negative emotions.
Ponder provides a safe, no-judgement space to recount our conflict. By asking clarifying questions, Ponder is able to understand the situation and then asks reflective questions that break the cycle of validation we might get from ourselves and others.
Though there are other therapy AIs that employ this framework of reflection, Ponder takes it a step further by utilizing storytelling. It is a generic educational strategy that is already well established and a way to promote a multiplicity of viewpoints in a digestible manner. It also develops imagination which is inherent in the empathetic activity of understanding different perspectives.
When individuals first use Ponder, onboarding will help Ponder understand the user and their personality, how they currently deal with conflicts, how they currently share/show emotion, etc. This helps Ponder as an AI cater the conversation to be most productive.
When individuals are using Ponder to talk about their conflict, Ponder picks up on keywords and tones in order to understand the conflict/situation beyond the user’s initial biases/emotions in order to ask better clarifying and reflective questions to the user.
The story that then is generated is based on anecdotes found through online articles, that introduces a perspective that the user may have not initially considered. The user is able to go deeper into the sources that Ponder pulls from if they wish to do so.
In order to and remember the lessons we learned through reflection, Ponder helps the user set goals to take small steps towards building a habit of empathizing.
Designing for AI: acknowledging boundaries
Be present when others can't
Introduce new perspectives for thought
Help you set goals for progress
Replace your human relationships
Force you to change your perspectives
Define & measure your "success"
By using Ponder, we hope to help individuals reevaluate their initial responses, think about their situations more thoroughly, and imagine different perspectives until a habit of continuous learning is developed where they would no longer have to rely on Ponder.
Love, team Ponder
Interested? Interested? Interested? Interested?