AI promises to track, understand and help us make sense of our bodies, minds and the world around us through thousands of datapoints. But how often is this experience delivered in a way that we can truly understand and benefit from? Through our work and research on pet wearables and smart home tech, we have seen the mismatch between consumer expectation of AI and what AI actually offers.
If we look at the landscape of AI and healthcare, we begin to see that AI is incredibly adept at tracking and collecting data, for example the vast realm of fitness wearables such as FitBit, and sleep trackers like Oura Ring. But often, these wearables fall short of giving solid recommendations, instead, they simply display data in beautiful ways. This is wonderful in that it makes the incomprehensible legible, but not so great in that it fails to give direct advice. As we interact more with these devices a sense of frustration and disillusionment is revealed. These devices are not as smart as we thought they might be. The burden of analysis still falls upon us. So, when can we begin to see how AI makes decisions and recommendations that can benefit us?
There are a few companies leading the charge. A good example is the health companion app Ada. Ada helps people self-diagnose different medical ailments through an intelligent chatbot and a vast database. What’s unique about Ada’s approach is that it uses an open-ended conversation format that allow users to answer honestly and feel heard. Much like how a real doctor might make a diagnosis, the conversation begins with an open-ended question that narrows down to specific yes, no or a little, a lot type questions. The app has excellent bedside manner and always allows users to ask “why” and provides detailed imagery and explanations when users have questions. The end result is often two to three different diagnoses with a clear explanation of what they are, what led to the diagnosis and what should be done next.
The transparency of Ada’s decision making, and it’s use of conversation as a tool land it into the new emerging field of Empathetic AI. Ada excels at making people feel seen and understood, allowing people to discuss their symptoms at their own pace. The intelligence and the delivery of knowledge is truly beginning to live up to the expectations and projections we see in science fiction. Empathetic AI embodies a fundamentally different mindset in designing intelligent yet human AI systems.
In the next article in our series on Empathetic AI, we will discuss what Empathetic AI is and how we expect it to change how we design experiences with AI.
This article was first published in July 2020