Just a few short years ago, there was a genuine fear that artificial intelligence (AI) and other automation technologies would soon replace workers at every level, from truck drivers to executives. Now, in the face of the Great Resignation and growing competition for skilled talent, AI is embraced rather thanfeared.

But before I explain how AI can help, it’s important to debunk the myth among executives that it’s all hype. In a recent study, KPMG U.S. found that three-quarters of C-suite executives said that the use of AI to help business was more hype than reality. Yet in the same survey, nearly 80 percent said AI was functional in their organization, and a majority said it was delivering more value than was promised or expected. There’s clearly lingering ambivalence and, in some cases, flat-out confusion.

While it’s true that AI currently has nowhere near the sophistication ofwhat you might see in a sci-fi movie — and frankly may notfor years to come — that doesn’t mean it’s still a nascent technology. We already have solutions that can perceive emotion. AI solutionscan learn. They can make sense of unstructured data. They can find and correct errorsthat are needles in a haystack. They excel at detecting fraud and predicting and quantifying risks. The list of capabilities is long, and getting longer by the day.AI doesn’t need to think like humans for it to have its own distinct and valuable function. We have a large toolbox, and the benefits are real.

Sohow exactlycan AI solvethe labor shortage? Can automation simply replace workers? Well, no, actually.

It was indeed true that in the wave of automation that swept through manufacturing in the 1980s, many blue-collar workers lost their jobs to robots. But by and large, the value those workers provided to the business began and ended with a single repetitive task — something the robots could do faster, better and more cheaply.

However, with today’s digital workforce, it’s exceedingly rare that any employee performs just a single task. Workers usually play multidimensional roles and move from one area of focus to another throughout the day.They make dozens of decisions each day. They collectively possess immeasurably valuable institutional knowledge that most businesses can’t afford to lose.

But even if you can’t use AI to replace workers, here is the real question executives should be asking today: “How can I use AI to improve employee productivity and satisfaction?”

Eliminating boring, repetitive tasks — ones that don’t require the knowledge, experienceor creativity of a human —can improve productivity and job satisfaction so that employees stay in their jobs. But it’s also important to enhanceemployees’ decision-making, which currently may be based on incomplete information or human-limited analysis.

For example, while AI may not replace a salesperson, a sales call can go better if AI is analyzing the customer’s emotions providing that information in real time to the salesperson.

In the recent case of one large retail bank, customers were emailing the bank with complex customer service questions. The bank actually received 4,000,000 emails a year, requiring approximately 600,000 man hours, and the resources simply weren’t there to handle it.  KPMG implemented a machine learning solution to triage issues and in some cases, automate communications so that their clients were being served in spite of the skills shortage.  It was over 90 percent accurate in determining customer intent.

Another financial services firm came to us with escalating workforce constraints related to customer care. As it moved to digital channels of communication and support, the need for personalized customer care was requiring unsustainable workforce additions. But the level of white glove service and the nature of the inquiries were too complex to simply outsource to an external call center provider. This company plans to install AI technology to enable them to rely on fewer people while still maintain its commitment for superior service to its customers by automating high call volume responses and triaging more complicated issues to appropriate subject matter professionals.

To address the shortage of skilled talent caused by the Great Resignation, you ultimately need to rethink your business. Simply doubling down on hiring and recruiting is a stopgap measure and not a long-term solution.

Instead, invest in AI solutions and start rethinking your workforce mix. Activities that require fewer skills have, for the most part, already been automated or offshored. But AI also has the potential to fill the gap for medium-skilled and back office workers who are currently in short supply. It arms employees with tools to do things better and focus on more high-value activities. All in all, it can augment human skills for complex tasks and increasingly take on simpler tasks.

AI tools abound. If executives take the time to invest in them and rethink their workforce mix, they will go a long way toward addressing their labor shortage.Our clients tell us they are so glad they did.

Brad Fisher is a senior leader at KPMG U.S. who specializes in implementing technology solutions to accelerate business results. He advises the C-suite and boards at Fortune 500 companies.