I have found that the tension between open form exploration, and the need to speak about realistic applications releases the most applicable value in my research, and broke it into new grounds.
Over the last week I found myself in one of these situations when I needed to suggest the consumer–facing applications of this work.
A couple of existing sentiments that I have had for some time but didn’t have much luck consolidating to a compelling alternative are the dismissal of anecdotal startups, design that is independent of a computable outcome and an eco–system that can accommodate more nuanced data models than the one–dimensional Model View Controller.
Thinking through the questions I was asked I realized that there was more to my fascination with single purpose algorithms.
Single purpose programs are independent of a domain.
We can think of digital products, say Facebook, FIN or Peach as islands. Digital destinations, built by their respective business, waiting for users to inhibit them through on-boarding bridges and move within using controllers, features and the guidance of marketing. These products, large or small, successful or forgotten, are driven by a core logic. It is on that logic that features are ideated, products are conceived, code is written and designs are based.
It is this gravity that makes a domain valuable for users. Traditional digital product thinking will make us believe that each logic needs a unique address. It needs a platform, hard–coded user avenues, and an incremental growth trajectory that fits in agile sprints.
Ironically the disruption to this model has come from the disrupters themselves.
In our new reality data is coming too fast, and from too many angles to fit in this paradigm. We collect, and are able to process useful data in volumes that one single domain can’t handle (with its binary data sets and fixed controllers). It is to this problem that we’re trying to deploy machine learning and AI. Using hyper–statistical computational processes as a mean to further the efficiency of our domain. Increased capacity, artificial decision making, general purpose agents and more.
Once we have established the language of domains, logic and single purpose environments we have everything we need to to design an alternative. What if we shift all of our focus to that single piece of logic? A great idea, and the disruption factor. Can we think of it as the predecessor to features, and products? Does this idea must be developed within the constraints we know? Does it need to be an app? Or a website? Or any other unique “place”?
What if it has no domain — what if views and interfaces were completely secondary to the value that logic could produce. No domains, no apps, no website, no text interface, no controllers.
Can we design next–gen products that interact directly with data?