The newest employees at businesses around the world don't show up to job interviews in freshly starched suits or conservative pumps. In fact, they don't have much in the way of a corporeal form at all.
We're talking about artificial intelligence by way of machine learning, complex algorithmic models capable of deriving valuable conclusions from data without manual intervention. Machine learning is the talk of the town these days, but for every organization that touts its utilization of such a force without really knowing what it is, another has integrated actual machine learning technology absent-mindedly.
Do businesses really understand the ramifications of starting down the path of machine learning? Or will Isaac Asimov have to crawl out from his grave and teach them a lesson on the Three Rules of Robotics? We think the famed author's first law in particular, as follows, sheds important light on this topic:
"A robot may not injure a human being or, through inaction, allow a human being to come to harm."
How can businesses integrating machine learning into IT operations make sure their robots stick to the script instead of turning into an army of unstoppable destruction?
Big Brother's Keeper
Commercial big data collection has certainly revved up in recent years, but as combing and conditioning data becomes a commonplace practice in the business sphere, consumer data gathering will get more precise and calculated.
However, machine learning technology is less discerning - it's hungry for data of any make and model, like a teenage millennial who eats quinoa and artisan bacon donuts in the same sitting. What businesses feed it matters to the health of the system, which will constantly push companies toward more exact data mining.
But as research from the Harvard Business Review shows, consumers are overwhelmingly ignorant about what sorts of personal data these businesses legally obtain from them. It won't stay that way forever. As people warm to the mild intrusion of ambient data collection, they will want to see some sort of return on their investment of information. Commercial machine learning technology must therefore always have at least an element of customer service, so long as customer data is what it's feeding on.
Take IT service management, for example. Machine learning technology in the latest products not only answers help desk and service desk tickets to prevent unnecessary manual intervention, but it converts the tickets themselves into resources it and users can utilize to further resolve issues and optimize continuously.
Finding a Place for the Atypical
Now that we've tackled the first half of Asimov's robo-rule, let's get to the second half: What sorts of damage could machine learning cause by doing nothing?
Machine learning occurs through something resembling an iterative process. Data of similar ilk enhances a program's ability to predict values, but at this stage of the science, that typically means within a narrow wheelhouse. As a result, machine learning may occur unevenly, like a student who pays attention in English class but dozes off during advanced trigonometry.
From a customer service standpoint, this can be incredibly infuriating, as machine learning may become more attuned to resolving common problems than uncommon ones. So, what happens then? Bad press. A Dimensional Research study commissioned by Zendesk in 2013 found one bad experience is enough to stop nearly 60 percent of customers from buying from a business altogether. Moreover, more people are apt to leave bad reviews on social media and widely share bad reviews with others.
So, what's the takeaway for machine learning-enabled businesses? When it comes to ensuring this burgeoning technology doesn't promote the perception of preferential treatment, do not be loath.