Sometimes the world feels like a completely random place.
Other times events converge in such a way that you decide maybe some vast conspiracy is at work.
The latest confluence of events started when I finally got around to watching last summer’s Terminator Genisys movie. It was what I expected: a battle between man and machine, with a newer origin story for the evil Skynet system that better reflects the technology we live with today.
My movie viewing would have remained a simple escape if it were not for a recent comment I overheard, citing concerns with the latest content analytics technology. Advanced analytics causes worry in some organizations because they fear it will create a system — similar to Skynet — that will know too much and take control.
Paranoia at its Best
As a whole I hope that this fear was simply an exaggeration on the part of those who shared the story.
If technology becomes advanced enough to achieve sentience and it then decides to wipe out the human race, its odds of success will not go down by our refusing to deploy it. We see enough system security breaches in the news to have no illusions as to the security of our systems for a determined hacker, much less for a sentient computer program.
Besides, unless your organization has the nuclear codes, why would Skynet care?
This fear may also be rooted in the increasing level of automation within organizations. The fear of losing control is very common and has inhibited the adoption of new technologies for decades. But reading what content analytics aims to accomplish as scary or a threat is misguided: Content analytics attempts to do what people have either failed to do or have done poorly in the past.
Content Analytics Solve Classification
The issue content analytics attempts to solve is this: organizations do not know what content they have.
Even if everyone implemented a perfect content management system (CMS) today, a large amount of content would remain that was created last week, month, year or decade ago that requires classification, management and potential deletion. Getting people to manage active information is hard enough, without having them apply current rules to large repositories of older content.
And let’s face it, no such thing as a perfect CMS exists. Business is defined by exceptions to every rule. That applies to processes and content. There will always be content that's hard to identify without applying some intelligence. Human intelligence is the go-to method, but the volumes to review are massive.
Content analytics, that same technology that makes e-discovery feasible through the use of predictive coding, can resolve this problem. Auto-classifying 50-70 percent of the content in an organization based on the processes surrounding the content is possible, but a chunk will always remain that cannot be automatically classified.
Take email. No process can auto-classify an email as most of it is written straight from the email client. A business can enforce rules to store all emails to the CEO and CFO in a Capstone approach — which beats asking people to classify their own email — but still misses emails and treats emails about lunch as important as those regarding the board of directors.
Now think of all the content on share drives or tossed into SharePoint, Dropbox and Box. What about all of the content on every computer?
Content analytics looks at that content, clears out duplicates, groups similar documents, and allows you to classify them. It can build a model of your organization's content that will allow businesses to identify new topics that may provide a competitive advantage.
Skynet Doesn't Want Your Break Room Donuts
So if Skynet is coming to wipe out humanity, it won't originate in the enterprise software market. It will result from embedding self-driving software in advanced robotics that learns how to behave by observing social media. Think Google’s software in Boston Dynamic’s robots being controlled by Microsoft’s Tay.
A little scary, but not the technology that most offices will deploy. Using analytics to classify content, identify trends and find answers is a long way from granting those same systems the ability to run a business (or wipe out humanity).
What this technology does do is free us from tedious work and helps us unearth relevant information so we can make the important decisions — the kinds of decisions we were hired to make in the first place, instead of figuring out where to store that email from Ann in legal.
Title image by Dmitry Ratushny