This blog is in no way attempting to reflect science fiction nor does it intend to instill fear within mere mortals. The following events are based in the reality of 2017 on true events that happened in 2016. The term “Non-Human Centered Design” is not a well-known term, it is merely the best description I could come up with for the following collection of thoughts.
In recent times, we have seen concepts like Human Centered Design take the world by storm. It is virtually impossible to find a start-up or Fintech company these days that does not have a story about how observations of human behavior led to aspects of a newer, slicker design. A classic case of for humans, by humans to humans – involving the human perspective in all steps of the problem-solving process. Many of the most heartwarming stories [http://unicefstories.org/tag/human-centered-design/] have one common thread – a distinctive human centered design. I, however, have had a different revelation.
Which was first, the chicken or the egg?
A few months ago, I was sitting in a design session designing the automation of a process that is currently extremely manual. This process was very mathematical by nature. At some point during the process, the subject matter expert on the team told us that we needed to round off the result to the nearest 0.5%. The justification for this was: “You would never say 6.6% you’d just say 6.5%”.
This single sentence led me to start thinking about Non-Human Centered Design. It was in this moment that I realised the mistake I had been making throughout my career. I realised that in some instances, you need to apply non-human centered design in your thinking. That means instead of focusing all your efforts on the users and their requirements, one could design the activity first, and the device will follow.
To be or not to be?
Non-Human Centered design is the term that I use for describing a thought process that I’ve recently started applying to my designs. It simply asks the question: “Does this rule/assumption/constraint only exist because of a human concession?”. In the situation described above, the automated process would not care if the number was 6.6% or 6.5%. This was a concession made purely due to a human preference for round numbers. With a non-human centered view, rounding off to the nearest 0.5% would be illogical in an automated process.
For the first time in history, humanity is beginning to design systems that are not made for humans to use. This is often seen in APIs and AI systems. I firmly believe that a Non-Human Centered approach will yield the best results. Who has paid for all of the “rounding off” that has been done in the past? In a price sensitive market, how many sales or deals have you lost because we’ve rounded off for human convenience. In today’s world we have algorithms making decisions, have we disadvantaged ourselves by making these human centered concessions where humans are not involved?
In a recent brainstorming session we were discussing possible uses for Natural Language Generation. It is now possible to get an AI system to produce detailed reports. There were multiple ideas that all had the same design pattern: ie. feed data to an AI system and a report will be generated. Applying a Non-Human Centered approach however made me question the relevance of the report itself.
The final call
We use data from our systems to generate these reports so that a human can make decisions and input actions into other systems. Many of the ideas discussed feature a future where a human is no longer needed to make the decisions. In this particular instance, the report itself is redundant, so why generate it at all?
In designing for the future, I think it is necessary to question what rules and constraints are in place merely because of human concessions. This is especially relevant when humans are no longer needed. Will humans or Non-Human Centred devices make the call?
Guest Blogger Tyrone Naidoo