Tech professionals almost reflexively associate artificial intelligence (A.I.) jobs with data, and with good reason. When Amazon announced plans to “up-skill” 100,000 American employees for a more tech-based workplace, the company said that over the last five years its fastest-growing jobs have been spawned by systems that gather, process and analyze data.
For example, the number of data mapping specialists Amazon employs increased by 832 percent during that time; the number of data scientists grew 505 percent, and the number of business analysts rose 160 percent.
But it’s also worth noting the dramatic growth in non-technology roles that have been touched, in one way or another, by A.I. Jobs such as logistics coordinator, process improvement manager and transportation specialist expanded by over 400 percent.
“A.I. creates a brand-new class of products that will require humans to build, maintain and deploy,” said Armen Berjikly, senior director of strategy for Weston, Fla.-based Ultimate Software. “A self-driving car will not make mechanics obsolete and will create a new class of programmers to boot. Analytics software will require a human to interpret, challenge and distribute results in the human world, and so on.”
A.I. Needs Infrastructure; Infrastructure Needs Management
For tech professionals, some of the best opportunities lie not in data science or the software development that enables it, but in the design, engineering and management of systems that support it. Technology isn’t the only challenge involved with implementing A.I. solutions, experts say: Budgets, a lack of skilled talent, uneducated executives and, increasingly, complications from tying together different functional systems are driving demand for new types of expertise.
“If you’ve got 5,000 employees and 20 data models for each one, and another several hundred for each department, and online you’re modeling all of the aspects of all of these things, you have to deal with Big Data models and somebody has to run that data model farm,” said John Sumser, a San Francisco Bay Area analyst who focuses on the intersection of people, technology and work.
Such jobs won’t be easy, Sumser continued. Since the amount of data being used by businesses continues to grow on all fronts, the data model farms he describes are sure to expand in size and complexity until they break. That means “you have to be developing alternative data models while the current ones are running, so when the tire goes flat you can put on a new tire,” he said. “At the same time, you have to be assessing the health and configuration of all those things.”
Mark Willaman agrees. The founder and CEO of advos, an Aptos, Calif.-based provider of marketing and sales software, Willaman recalls speaking with a Fortune 500 manager about how their company’s tech stack just keeps growing: “Five years ago, they used five products to find talent. Now they use 40.”
Beyond Data: Integrating “Smart Stacks”
Business executives, Willaman points out, usually think in terms of problems to solve rather than technology to buy. That’s why in many cases, what on the surface seems to be an A.I. or data science position may really be about identifying tools that should be pieced together to “get the job done that we’re trying to do,” he said.
“That’s really complex these days because not only are the number of products increasing, but within each product category there’s dozens to choose from,” he observed. Tech pros who can step back and understand business problems, review A.I. tools in their context and put together a workable suite are going to be in demand. Accomplishing that is “part art, part science and part business acumen,” Willaman said.
For example, organizations need interface designers who can present the analysis of data. And as they gather more types of data to generate more types of intelligence, they need to find tech pros who can determine the most efficient ways to gather it, store it and scale everything they do.
Even when they live on the best of infrastructures, data and A.I. capabilities do little good if employees and customers can’t take advantage of them. That’s why more solutions providers are producing technologies to make complex information and sophisticated computing capabilities more accessible. For example, said Omer Biran, SAP’s head of conversational A.I., customers want “‘conversational UX experts’ who understand their business and can ‘conversationalize’ the interactions users have with services and products the customer creates for their [own] users.”
And, Berjikly adds, many of those conversations won’t be about data or computer systems. “With the A.I. and advanced technologies of our era, we are effectively creating a new approach to things we do at work and leisure,” he said. “Whether it’s manifested as robots or something as ubiquitous as cars … [we] will need vast networks of repair and maintenance facilities and related staff. Software that prioritizes customer feedback will need talented analysts to interpret and potentially argue with its recommendations,” he said.