User-Centric Design & Machine Learning

Is the process of creating an empathetic experience inherently human? Or can machines help us get there as well?

Written by
Andrew Frank
August 2, 2017
Filed under:
Data Science

Empathy is The Best

An old colleague of mine recently posted an Instagram image of a presentation slide. The caption on the slide reads: “Without empathy, it’s not design. It’s fine art.” I was taken aback by this; it just didn’t feel right. So I commented on the post:

I’d say it’s more along the lines of, Without empathy, there are only passive experiences. Art by definition is subjective and meant to provoke a reaction based on your own previous experiences. As a self-appointed representative for the UX & Product Design community [for this comment], our interpretations of others’ experiences, emotions and biases need to skew towards the objective spectrum in order to leave any personal biases out of the equation. We can make presumptions based on what we learn, but to draw a line between fine art and design based on an opinion without a user-centric context is invalid in my opinion.

So, this became more of a rumination about what empathy brings to the table rather than what fine art is. (I had plenty of these conversations when I was 20 years old and in art school.)

Art is subjective, even if its intent is to critique objectivity. It’s open to interpretation. Empathy, however is not open to interpretation. Empathy is the act of embracing another’s customs, culture, emotions, experiences, and behavior to experience through their unique perspective. You must leave yourself out of this process to achieve the depth of empathy that creates a viable user-centric design solution.

User-Centric Design & Machine Learning

This got me thinking about disciplines beyond art and how they relate to design theory and the creation of great UX and user-centric products. Are we really doing our due diligence by making presumptions based on observations when we have our own personal biases? Or is there a more empirical way of achieving the level of empathy necessary to solve a problem?

It comes down to a question: can behavior change be achieved when any sort of benchmark for user behavior [empathy] has been established? After all, that’s our ultimate goal as we’re guiding users through any experience with a series of strategic interactions.

Machine Learning as an Empathy Machine

I would like to propose that the collection and analysis of a user's data over time qualify as a legitimate strategy for establishing benchmarks when identifying any user's unique behavior. Following this train of thought, we can assume that it’s possible to create unique experiences comprised of targeted interactions and user flows — or suggestive content can be offered up as a machine learned version of an empathetic experience. Of course, this leaves the human element of interpretation out of the equation, but as stated before, empathy in its purest form should be as objective as possible; in this rumination, there’s no wiggle room for personal biases.

User-Centric Design & Machine Learning

So if machine learning can indeed achieve this theoretical baseline benchmark for a user's behavior, and return a personalized empathetic experience, we — by definition — have reached a level of empathetic understanding through technology. It’s a quantitative empathy that can uncover other qualitative traits by examining user behavior over time.

Problem Solving = Achieving Behavior Change

All design is a philosophical methodology for solving a set of issues that comprise a larger problem. Design combines ethnography, anthropology, creativity, personal intuition, and a logical and methodical approach of discovery to come up with a solution to a problem. The scientific method does that as well, albeit under different conditions: hypotheses are presumptions until proven true or false, and presumptions are inferences based in fact (previous discovery). The only difference is that design methodology relies on achieving empathy to induce behavioral change. I believe the same principle applies to using technology as a tool for discovering and achieving empathy. One is never going to be empirically better than the other; they both have their uses.

User-Centric Design & Machine Learning

So when it comes to achieving empathy, I believe that there is no set of processes, tools, or questions that are going to be more effective than another strategy. Human-learned empathy certainly can be effective, as it involves a level of personal intuition, but that intuition skews toward a personal bias. A machine learning methodology may take longer — it must spend the time to watch and analyze over a longer period of time — but it’s completely void of biases and subjectivity.

There is No Spoon

Ethnography, the act of discovering another’s personal culture and customs, is inherently set up to do this. It starts out with a layer of objectivity, which sparks the interviewee’s self-analysis. This is where UX truly starts: at the conversational level, achieving a thorough understanding of others' experiences, behaviors, emotions, wants and needs. As we begin to understand and analyze the data — and inject ourselves into the stories we’re told — we must always keep in mind the conditions and behaviors we’ve uncovered. Empathy is our tool, our approach to do this successfully.

If you got The Matrix reference back at the header of this section, go ahead and give yourself a pat on the back. If you’re not familiar: it’s a sci-fi movie about a dystopian world run by robots, where humans live in a computer generated reality. Neo, the main protagonist and hero of the trilogy, learns he has the power to alter the computer-based reality from a young child who is bending a spoon with his mind.

User-Centric Design & Machine Learning

The explanation posed for this phenomenon: both Neo and the child are aware of their existence within a virtual reality. Even though there is no physical material, and the spoon does not physically exist, anything and everything can still be done to that spoon without any of our real world limitations because a spoon is an idea. Material does not need to exist for an idea to exist; but without context, there is nothing.

Without an idea or concept, there is no problem to solve. Without empathy, there is no user-centric design. Without user-centric design, there is no basis for benchmarking behavior and response to stimuli. Any method that obtains the necessary level of empathy to solve a user-centric problem through behavior change can be validated.

Of course, that’s our opinion. What’s yours? Is empathy truly exclusive to humans — or can it be achieved via other means?