More businesses than ever rely on artificial intelligence to either help employees make correct decisions, or automate certain processes entirely.
Those involved in the development of AI believe that, as the technology’s capabilities grow, so will the need for architects, engineers, developers and data scientists who can put it to practical use. Even so, at this early stage, the idea of “automating” judgment so that systems can perform without any human oversight is a bit premature.
“There’s always a person involved,” observed Ellen J. Bass, Ph.D., head of the Department of Information Science at the Drexel University College of Computing and Informatics in Philadelphia. “Think of a deep spacecraft, which is highly automated, but also in regular touch with flight controllers. It’s very rare that you’d have a complete, lights-out system. And even if you did, remember that some person designed it.”
But with the Internet of Things (IoT) beginning to connect more devices in new ways, the technology’s growth could prove dramatic. In a recent article on the rise of intelligence systems and automated judgment, research firm Gartner described a farming system that can track weather, soil moisture, acidity and other factors, and use that data to decide where and when to plant a new crop, even using GPS to properly align the field.
Skynet and IoT
The Internet of Things could make automated judgment truly ubiquitous. “In every industry around the globe, someone is thinking about warranty costs a different way, adding a new service to run or manage a device that once was just sold or dreaming up a new supplier network or an alternative way to distribute products to new direct buyers,” wrote Chris O’Connor, the Raleigh, N.C.-based general manager of IBM’s Internet of Things offerings, in a recent blog post.
“Judgment implies ambiguity,” O’Connor told Dice Insights in a separate interview. “You’re bringing information together, but somewhere in there is a discussion around, say, risk tolerance, and judgment has to be applied.” More often than not, he believes, that judgment involves human interaction. As sophisticated as a system might be, “humans can see things that aren’t always obvious. So, there’s a person in there validating the system’s judgment.”
“I think Gartner’s view [of AI] is very business intelligence-centric,” added Patrick Dowling, vice president of business development for LogicNets, a Washington, D.C., provider of decision-support systems. His customers manage complex knowledge environments in areas such as technology and healthcare. Rather than depend wholly on a system to make decisions for them, LogicNets users rely on software to analyze information and walk them through solving daily issues in areas like tech support or health and safety.
Of course, LogicNets isn’t the only organization at work on such applications. Specialized firms like Strata Decision Technology and Inovalon, as well as giants like IBM and Salesforce, offer some flavor of decision-support systems. Some of these, like LogicNets, are designed to help users navigate through multiple information paths. Others, like IBM’s Watson, push beyond the idea of streamlining systems or even medical diagnostics; instead, they’re combing through large amounts of data with the aim of identifying solutions that no one has ever thought of before.
Foresight and Fundamentals
“Intelligence is where the rush and opportunities are today,” said IBM’s O’Connor. “Taking a bucket of bits and figuring out what to do with it. That’s where the industry demand is.”
With that in mind, O’Connor foresees growing opportunities for professionals who can mix their technical knowledge with an understanding of information and data workflows and specific industry expertise. “This is about taking interconnected things and making them smart,” he explained, adding that the best prospects will come to people who recognize areas where smart systems can lay the foundation for innovations that transform processes and businesses.
That means that all tech pros should be paying more attention to data, he added. For example, O’Connor suggests that data center personnel could create transformative ways to set up and run their operations by studying the data they generate. People with specialized knowledge “can challenge a data scientist to make it happen,” he said.
For all of these systems’ advanced capabilities, however, the people who design and build them need to pay attention to the basics. “Think about the privacy and security side,” Bass said. “How do you make sure airborne cameras are safe and not hackable?” If you’re going to trust a system’s judgment, you have to make sure it’s secure and can’t be taken over.