Much of the conversation around AI at work has focused on a seemingly straightforward question: How much more productive can we be?
Alexis Krivkovich thinks leaders should be asking a bigger one: How should we redesign the organization?
Krivkovich, who coleads McKinsey & Company’s People & Organization Practice, estimates that 57 percent of today’s work activity can be done with existing AI technology. She is quick to clarify, however, that her estimate does not mean half of today’s jobs are on the verge of disappearing. Instead, Krivkovich sees a future where jobs transform and judgement and relationships become more important than deep expertise on a single subject.
“If most of the roles need to be reimagined, who’s doing that?” Krivkovich asks. “How do you make sure you’re not leaving all of that to every individual in your organization to figure out on their own?”
In her conversation with The Atlantic CEO Nicholas Thompson for Off Site, Krivkovich explores what these changes could mean for how companies develop their people and structure their organizations. AI could allow employees to redefine their jobs for the better so they are free to focus on higher-level activities. But it also creates new risks. If AI performs the routine work through which young employees once developed their skills, how will they build the judgment and the relationships they’ll increasingly need?
Krivkovich sees an opportunity to make human development a bigger part of the job. By removing some of the day-to-day drudgeries that make up so much of a person’s workday, AI could create more time for coaching, advising, apprenticeship, and the professional relationships people have long struggled to prioritize.
“We’re going to need more of that, not less,” she says.
All of this makes AI adoption as much a leadership challenge as a technology challenge. Krivkovich says companies whose employees have greater trust in AI aren’t simply distributing tools. They’re redesigning roles, investing in upskilling, changing incentives, and having senior leaders demonstrate how they use AI, including where they get it wrong.
The transformation, in other words, may be technological in origin, but Krivkovich believes its success will depend on people.
In her conversation with Thompson, Krivkovich explores what work might look like when companies stop asking how humans can adapt to AI and start asking how work itself should adapt.