Machine Learning to rethink the Human Experience

As Artificial intelligence converges with User Experience, where does that leave design?

alt text

Will UX become a component in AI or an equal player? To be sure, the access to machine learning UX platforms poses a challenge. Working with emerging technologies must be rethought beyond form and function. If you are designing in an an environment where machine learning is beginning to be implemented, read on. Here’s an approach to consider to help you spot product, service and system opportunities.

First, start with the right ground work. This framework assumes designers are versed in at least the basics of weak AI. Your mindset must also take into account that the gap between input and output is disappearing, and you must have an increased awareness of how the physical world is interconnected as never before and how user friction is being reduced.

We are done with just developing mobile apps and web applications. We are now designing services, within a machine learning environment. The Organic Machine Learning Design framework identifies three starting points for designers to contribute newer, stronger AI possibilities:

Biological Plausibility - Does it work in a real world setting?

Mimics how the brain learns with biological subjects and settings. Examples include, advances in face recognition development which exploit naturally occurring differences in faces and form the basis for system perception and forecasting of data. New information is acquired through “associative learning”, in other words via constant feedback loops between stimulus, response and representations of events. Representative of conditions in the wider world and population, provides the ability for ecological validation.

Incidental Learning - are you expecting a collision of outputs for a simple input of data?

Content, media, and data are already broken up into “chunks” that form part of a larger “whole” or system, the degree to which these collide make way to discover new ways of learning and information. Not premeditated settings but acquired because of another process, not the case anymore where an input is associated with a specific output as is the case of traditional associative learning. Leads to innovation in designs where recognition and comprehension systems play a critical role.

Force for Motivation - Does it motivate improvements and make humans better?

Technologies exist where a computer can administer clinical tests on humans, taking over assessment and forecasting tasks. Due to ability to process data they are faster than humans at responding to comments, questions and nonverbal cues. However, this is speculative and controversial as professional communities are faced with a full digital transformation to accept this reality.

How Machine Learning brings User Experience Value

Some things to consider when designing with machine learning to enhance user experiences include good olde’ usability heuristics like user control. Currently the design of human experiences with Artificial Intelligence painfully falls short when systems attempt to anticipate needs and pain points by responding while lacking true contextual awareness. Consider giving the user the impression that they are in control and not totally giving it away to an artificial intelligence system. Otherwise the risks may include system responses that are plainly out of context or potentially offensive, leading to user annoyance or product abandonment. A safe way to handle unpredictability is to always work towards a meaningful goal for the user. Simply put, the value does the innovation brings to the human experience beyond mechanical automation of turning appliances on and off.

Another way to handle this ambiguity is to turn off artificial intelligence and provide the user the ability to take over manually, or override settings if the system is about to make dangerous or compromising actions that can affect users. Scaling back is ok. Scaling back artificial intelligence features doesn’t mean disappointment especially if users are not adopting or using the service, system or product as expected. Any effort in the field of user experience that applies artificial intelligence theory must always emphasize the meaning and value brought to the human, since the goal is precisely this delivery as a service especially when the context of use matters.