Gathered at Fortune‘s Brainstorm Tech conference on Wednesday, a panel of leaders from companies including Lenovo and McKinsey argued that, when used with proper controls, AI-powered tools can free up workers to focus on more “human” tasks.

“It takes toil out of our work so we can increase joy,” said Lareina Yee, a senior partner and technology council chair. “It can create a massive amount of productivity for the individual.”

A technology revolution

Ever since OpenAI released ChatGPT to the public at the end of 2022, executives and tech critics have clashed over the promise of AI—and the risks to workers. While economic leaders such as the San Francisco Reserve chief Mary Daly argue that AI “replaces tasks, not people,” the fear remains of massive layoffs and replacement. The World Economic Forum has estimated that AI will replace around 85 million jobs by 2025.

Speaking on Wednesday, the panel of business leaders took on an opposite stance, arguing that AI will lead to unprecedented productivity gains for knowledge workers. Vijay Gopal, an executive of digital workplace solutions at Lenovo, described AI as the “single opportunity we have as mankind to achieve perfection.”

AI can take over tasks from summarizing meetings to updating legacy code, as well as assisting workers with modern programming languages like Python—a role that Yee said could give 20% of the workday back to employees. “How would you use that time differently?” she asked. “This is the first major technology shift that actually addresses knowledge work.”

Athena Karp, a general manager at Workday, said that AI could go even further by helping to remove workplace problems for employees such as bias. She pointed to the example of hiring, where an AI tool scanning resumes could remove biases inherent in humans, such as prioritizing applicants who went to the same school or grew up in the same hometown. “You can’t govern bias in humans, but you can in technology,” she said.

With the cycle of AI adoption still in its infancy, Karp said there is a massive opportunity for workplaces to rethink their internal processes—as well as implement adequate controls and governance. She added that 80% of recruiting organizations that she works with do not yet have an AI-first product in place.

For Ali Akhtar, the cofounder and CEO of the software startup Letter AI, the technology is already here. He said that when conducting job interviews for software engineers, he gives bonus points if the developers use their favorite AI copilot, illustrating that they’re already thinking about how to leverage tools to increase their productivity.

While concerns remain about the errors still inherent to generative AI products, Akhtar compared the mindset to self-driving cars, where products have to perform 10 or 100 times better than humans to reach acceptance. “How much better does AI have to be so that we’re forgiving of any mistakes it might make,” he asked, adding that the use cases will expand as large language models continue to improve.