AI will reshape the global labor force. Employers will need to help their workers keep up.

This article is part of " Workforce Innovation ," a series exploring the forces shaping enterprise transformation.

After a full day of seeing patients, Missy Scalise used to spend hours at home fleshing out the notes that go into medical records.

Then, about a year ago, Scalise, an internal-medicine physician with Ascension Saint Thomas in Nashville, started using Suki , an artificial-intelligence tool that generates notes and other data based on her patient conversations.

"What used to take me two to four hours per full day of clinic now takes me about 30 minutes or so, at most," she told Business Insider.

Having AI working in the background to document what goes on in the exam room has given her more leeway to focus on patients, Scalise said. Several have told her that she seems more present, she added.

Part of Scalise's role is chairing a committee on clinician well-being. A couple of years ago, AI wasn't on her radar. Now, she's pushing to get more colleagues to start using the technology to help boost their efficiency.

That highlights an important reality: AI will likely change some aspects of nearly all jobs. To keep up, workers will have to add skills and learn as never before, and that will require employers' help.

The breadth of the change to come seems profound. In January, the International Monetary Fund predicted that AI could affect nearly four in 10 jobs globally. For advanced economies, the share of jobs exposed to AI rose to about six in 10 — about half of the latter in a negative way.

In 2023, the World Economic Forum reported that employers expected 44% of workers' skills to be "disrupted" within five years.

It all points to a workforce that will need widespread retraining .

Ravin Jesuthasan, a coauthor of "The Skills-Powered Organization" and the global leader for transformation services at the consulting firm Mercer, told BI that chief human-resources officers and other leaders would need to think of training — particularly around AI — as something that's just as important as, for example, building a factory.

"Everyone needs to be really facile with AI," he said. "It's a nonnegotiable because every piece of work is going to be affected."

Jesuthasan said that when he talks about AI's widespread reach, he often gets questions like, "Well, what about the plumber who came to my house?"

Plumbers, he said, might spend 30 minutes of every few hours turning a wrench. Much of the remainder goes to things like scheduling appointments and understanding the setups of customers' fixtures and piping. AI can help with all that, Jesuthasan said.

He said experimenting with AI was a good start but not a viable long-term strategy. More organizations are becoming deliberate in how they invest, he added. That might look like identifying well-defined areas where they will deploy AI so that everyone involved uses the technology.

That's part of a mindset of continual learning and reinvention. Without that, he said, both the organization and its workers risk falling behind.

Ultimately, Jesuthasan said, organizations should test and learn so that they can be judicious in how they deploy AI and see a return. This shift will require that leaders focus more on skills and less on the strictures of a particular job.

It's a big shift.

"We're talking about the reversal of 140 years of learned behavior and organizational structure and processes to shift from jobs to skills as the currency for work," he said. "You can't eat the elephant in one bite."

Jon Lester, the vice president of HR technology, data, and AI at IBM, told BI that business leaders should give experts in various parts of their organizations the time and space to learn what's possible — and to fail. He also said leaders should track the return on investment of their efforts and let their expert employees make decisions.

Lester said it's insufficient to drop in the technology without giving other considerations.

"If we gave this technology without the learning, without the focus on process and experience, we just automated bad processes," he said.

Another challenge with inserting AI into an organization is that it's often difficult to tell which skills workers possess and where they might need training.

Asteri AI is a startup that uses large language models to determine the skill levels of an employee base, identify gaps, and forecast what abilities might one day be needed.

Julia Grace Samoylenko, Asteri's founder and CEO, told BI that skills are often hidden and difficult to capture and that they tend to change rapidly.

"We truly believe that AI is the key to solving this problem," she said.

It's important for organizations to understand the capabilities of their workers , in part, because it's not advisable — or even possible, in many cases — for employers to hire their way out of a skills deficit, Samoylenko said. That can be especially the case with high-demand areas like AI.

"We strongly believe in the idea of investing in internal talent rather than bringing in external hires," she said.

Aside from concerns about maintaining morale and institutional knowledge, training existing employees can also be easier on the budget. That's because the cost of replacing an employee is, conservatively, half to two times the worker's annual salary, Gallup found.

Major employers like Johnson & Johnson are also using what's sometimes called " skills inference " to get a snapshot of workers' skills and how they might be improved.

For employers that want to train their workers , an obvious challenge is making sure it's effective.

Elise Smith, a cofounder and the CEO of Praxis Labs, told BI that AI could make learning more personalized. The company uses the technology to simulate various work situations. That might involve giving difficult feedback or managing someone who's being disruptive to a team, she said.

Praxis Labs is focused on middle managers and potential strong performers because, Smith said, training on these types of interpersonal skills is often concentrated among C-suite leaders.

"They often have executive coaches," she said. "They're doing role-plays. They're getting coaching on how to navigate these moments all the time." By using AI, Smith said, that type of training can be scaled to reach more workers.

Managers can give oral or written feedback in simulations, and the AI can interject to point out slipups.

"Midconversation, if you haven't asked an open-ended question, we might nudge you to do that because you may not fully understand the situation," she said. "And asking questions is a powerful skill."

Smith said the benefits of training could extend beyond the knowledge itself. She said clients had reported managers and employees feeling more engaged. And if the training is successful, it's more likely to stick, Smith added.

"You're more likely to change your behavior when you're confident about what you've just practiced and that skill you've just built," she said.

Organizations are also looking to use AI to empower workers in other ways.

Nathalie Scardino, Salesforce's chief people officer, told BI that the company's philosophy is that having AI tools at hand helps workers learn and improve their skills because the tech becomes incorporated into their daily workflow.

"You're bringing them along in the continued learning experience," she said.

Salesforce has pumped out more than 50 AI applications to help its workers do their jobs in new ways, Scardino said. The goal is to make sure workers continue to learn and experiment.

"Everyone should be able to work with AI and understand the use cases behind the advantages of working with AI," she said.

Charlotte Relyea, a senior partner at McKinsey & Co., believes that leaders need to plan now so they can better forecast the abilities workers will need. Given the importance of this task, she said, the process will likely involve not only the head of HR but also the CEO.

"It's going to require real investment in upskilling and reskilling," she told BI. "So organizations need to be prepared for that."

Relyea said leaders should prioritize where they focus most of their efforts on building skills. So it's probably more worth it to train a team with low turnover than one that's always seeing new faces.

Leaders need to think about this as a transformation — what she described as "painting the picture of the future."

In a recent McKinsey survey , nine in 10 workers said AI tools could improve their work experience. And, often, workers reported incorporating AI into their work faster than their employers.

"How do you actually harness some of the energy that employees are putting into using gen AI?" Relyea said.

She said employers should consider how they could create ways for workers to share ideas and make testing and learning part of the culture.

Scalise, the doctor in Nashville, believes that AI tools like the one she's using will be the future of medicine. She's pushing to give the internal-medicine residents she oversees access to the technology as well so that they don't get burned out by administrative tasks and can better focus on patients' needs. She said in the 18 years she's been a doctor, no other tool has changed her practice to this degree.

"I didn't know it would be this groundbreaking," Scalise said. "I'm ecstatic that it's been so effective."