Executives don’t see many job cuts ahead a result of tasks being replaced by AI. Is this a realistic perception?

A recent survey of executives out of IFS tackled issues of AI perception, finding few business leaders predict worker displacement by AI. Close to half, 46%, predict AI will actually increase headcounts over the coming decade, while 25% predict no changes at all to workforce sizes. Only 18% see AI as a tool for replacing workers.

There are many high hopes for AI — 61% see it boosting the productivity of their workforces. Another 48% also see AI as adding value to their products and services. While a majority anticipate productivity increases from AI, only 29% say such increased productivity will reduce headcounts over the next 10 years. “Respondents did not make the connection between increased productivity and reduced head count,” the report’s authors suggest.

If there are job losses, it’s likely to be in larger companies, the survey finds.

Areas where AI investments will be heaviest include the following:

• Industrial automation/machine learning (approximately 64% of large companies, 40% of smallest companies)
• Inventory planning/logistics (40% of largest and smallest companies)
• Production scheduling (38% of largest companies, 34% of smallest companies)
• Customer relationship management/sales (30% of largest companies, 25% of smallest companies)
• Service scheduling (28% of largest companies, 24% of smallest companies)

How to prepare workers for the changes ahead? More education, executives urge. A majority, 56%, state that society could best prepare by changing educational programs to prepare workers to make direct use of AI tools to increase their own productivity. Another 23% say they expect the market to create new jobs for people displaced by AI, while 15% even suggest a shortened 30-hour work week.

New products and services are the next “real frontier” for AI, the survey’s authors predict. As Bob De Caux, VP of IFS put it: “You can have increased productivity without reduction in headcount, particularly if you are using AI for augmentation instead of replacement. But you really have to get it right. In some industries and categories, productivity may be expressed less in units sold and more in speed to market, the number of design improvements and iterations, and the ability to give consumers more current technology faster.”

Reducing headcount for efficiency’s sake is not the goal. AI may boost the ability to delivering superior services and products, and this may actually require more skilled people than less.