In contrast to technical visionaries such as Bill Gates, Elon Musk, and Stephen Hawking -- who worry that artificial intelligence could become so advanced that it threatens humanity -- business leaders largely see AI as a way to enhance human endeavor.
Narrative Science, a company that sells AI to businesses, published its "2015 State of Artificial Intelligence & Big Data in the Enterprise Report" today. The report explores the attitudes of 200 respondents, including executives, data scientists, and managers, about the impact of AI on the enterprise world. For these professionals, AI is an opportunity than rather than a threat.
"Eighty percent indicated that AI makes workers more productive and creates jobs," the report says. "As opposed to replacing workers, organizations are leveraging AI for machine learning, automated customer communications, data-driven reporting, and business-decision making."
The report looks mainly at how enterprises are using AI to make sense of vast stores of data. More than half (58%) of respondents said they already have some form of AI such as machine learning, virtual personal assistants, decision support systems, automated reporting systems, and robotics.
[ Does automation really matter to IT? Read Driverless Cars, AI, Robots: Why CIOs Should Care. ]
"Given that we live in a world of headline and hype, it's easier to say that everyone is worried about AI killing us and taking our jobs," said Kris Hammond, Narrative Science cofounder and chief scientist, in a phone interview. "But almost everyone believes that AI is going to create more jobs."
Almost everyone among the business executives surveyed -- a group that Hammond concedes may have some bias as beneficiaries of AI-driven business processes -- may believe in the benevolence of AI. But there are also those who speculate that AI will eliminate jobs. A 2013 report from Oxford Martin School's Program on the Impacts of Future Technology suggests that "about 47% of total US employment is at risk" from automation. In January, Andrew Ng, chief scientist at Baidu, expressed similar concerns about the potential for AI to eliminate jobs.
AI will both create and eliminate jobs, as technology has done in general throughout the industrial era. What remains uncertain is whether there will be more creation than destruction, and whether those left without jobs will be able to adapt to the demands of new positions.
Hammond observed that factory automation has been eliminating certain kinds of jobs for decades. "Lifting heavy things and positioning things in exact locations turn out to be something machines do phenomenally well," he said.
It's different among knowledge workers, Hammond suggested, where jobs consist of many tasks and the automation of a single task frees the worker to focus on other responsibilities. To date, he said, AI systems have been aimed at specific tasks, rather than whole jobs. And he expects that will continue.
"If you have IBM's Watson providing medical diagnostics, it's not doing the entire task of being a physician," said Hammond. "It's helping with one aspect of that."
On Wednesday, Gartner issued a related report, "Smarter Machines Will Challenge the Human Desire for Control."
"As smart machines become more capable (and more affordable), there is no doubt that they will be more widely deployed in multiple roles across many industries, replacing some human workers," the report says.
Yet, Gartner's report also predicts that new industries and jobs will emerge. Rather than second-guessing whether there will be a net job gain or loss, Gartner analysts Stephen Prentice, Dale Kutnick, and Tom Austin argue that CIOs need to educate others; evaluate the risks and rewards of increased machine-driven decision-making and unrestrained data gathering; and stem potentially damaging misinformation about unstoppable machines that may emerge as AI plays a larger role in companies.
To put it another way: The goose that lays the golden eggs depends on artificial intelligence and must be chained and pampered at the same time. It's a thankless job but someone's got to do it, at least until management can figure out how to automate the position.