PwC Report Finds Workers Concerns About Losing Jobs to AI Overrated

Will artificial intelligence really take away our jobs? How will it resolve current big data dilemmas? Should organizations invest in machine learning to thwart cyber-hackers? These and other critical enterprise questions are addressed in a recent report from PwC, titled “2018 AI Predictions: 8 Insights to Shape Business Strategy.” In addition to technology shifts, the predictions explore the potential societal impact of AI. As a result, organizations may need to open their AI “black box” to validate the models behind their deployments, especially when they lead to rejected mortgage applications and significant financial market shifts, among other possible outcomes. This slide show presents highlights from the report, which includes research compiled from a number of PwC surveys.

PwC forecasts that the share of jobs at potential high risk of loss due to automation is only 1 percent by 2020. Two-thirds of executives believe organizations will benefit from people and machines working together to boost both artificial and human intelligence, but this will require a reskilling of today’s employees.

More than seven of 10 executives believe AI will deliver business advantages in the future. But what about today? While the creation of new industries is a decade off, enterprises can deploy AI now to automate processes that are too complex for older tech, identify trends to create business value and provide forward-looking intelligence to strengthen human decision-making. PwC reports that 54 percent of execs said AI is already increasing productivity.

Nearly three of five execs feel that the use of AI would improve big data initiatives at their company. Benefits should only grow over time as “lean” and “augmented” data learning techniques take hold—you won’t need to collect data from an entire fleet of autonomous cars to assess how they’ll drive, for example. You’d just need a few cars and a sophisticated AI math program.

Two-thirds of jobs requiring data science and analytics skills are in fields other than AI. “Citizen” data scientists must understand tech “basics,” but then know how to prepare and contextualize data so AI can leverage it for enterprise success.

More than one-quarter (27 percent) of executives said their organization plans to invest this year in cyber-security tools that take advantage of AI and machine learning, according to PwC. More will need to do so, as AI enables hackers to “follow” employees’ behaviors on social media, and then customize phishing tweets or emails to trick them into a network compromise.

Within time, AI will be capable of evaluating mortgages, flagging suspicious individuals at airports and making big bets on the stock market. Both businesses and the general public will demand to know how machines are arriving at such decisions, which will require more transparency from organizations—with proof of mathematical certainty—about the validity of their AI models.

The global AI “pie” will amount to $15.7 trillion by 2030. To establish the U.S. as a powerhouse, leaders are calling for increased federal funding, regulatory changes, standard/benchmark definitions, workforce development, and the launching of shared public data sets and environments.

New tech brings new fears, as 77 percent of CEOs expect AI and automation to increase vulnerabilities. Public-private partnerships are needed so citizens can collaborate with industry on AI’s societal impact. Self-regulatory organizations can establish agreed-upon AI principles for companies, and then oversee and regulate compliance.