Generative AI may be the first technological advance to allow for greater automation of service and knowledge work, whether it’s at a call center or a management consultancy. But does letting workers generate emails or PowerPoint presentations faster really lead to greater productivity? Ramine Tinati, the lead at Accenture’s APAC Center for Advanced AI, speaking at the Fortune Brainstorm AI Singapore conference last week, wasn’t so sure.

“If you give employees a tool to do things faster, they do it faster. But are they more productive? Probably not, because they do it faster and then go for coffee breaks,” Tinati explained. 

Instead, “if you reinvent the work, then suddenly those coffee breaks don’t become meaningful anymore, because you’re doing something else,” Tinati said, adding that some companies in Asia may be slower to adopt AI because “they don’t think about reinventing the work.” (Accenture is a founding partner of Brainstorm AI.)

Companies have, of course, been embracing forms of artificial intelligence to boost productivity for years, even before the release of ChatGPT in late 2022. May Yap, chief information officer at manufacturing solutions provider Jabil, said that her company had been using automation and AI to augment their so-called Golden Eye, the army of workers inspecting phones for scratches and blemishes.

Golden Eye workers spend eight hours a day on inspections, and working that long means that “errors will creep in,” Yap said. AI helped to augment the inspection process to account for possible mistakes from human workers. 

Chee Wee Ang, the chief AI officer at Singapore’s Home Team Science and Tech Agency, a government agency that develops tech capabilities for national security, said AI has helped improve processes significantly.

“Some of the information extraction … we see like 200% [improvement]. So that’s a significant improvement in terms of ROI,” Ang said.

Yet Ang also pointed out that beyond improving productivity, AI advancements are allowing Singapore’s Home Team to do things that it couldn’t do before, like responding to new kinds of crime or emergency situations. Singapore’s Home Team has 10 departments, including the police force, emergency services, and immigration authorities.

Reskilling

AI will inevitably lead to some job losses as certain roles become obsolete. And that can unnerve employees who are worried about getting automated out of a job. Employees already report concerns that they are being used to train their AI replacements. 

Panelists last week agreed that the way forward for affected employees would be reskilling and moving people into adjacent roles. 

“Transformation is scary, right? When you hear the word ‘transformation,’ people don’t like it,” Yap, from Jabil, said last week. She made it clear that Jabil wanted to augment, not replace, its human workforce. She added that “general skills sets” and “good leadership traits” cannot be taken away by AI, regardless of how it might automate other tasks.

Ang added that it was “very difficult to find in Singapore [those] familiar with [generative AI],” meaning that his team has hired people with adjacent skill sets but without direct experience. Another limitation? The lack of GPUs, as the Home Team has to work with on-site processors owing to the sensitive nature of its work. 

Tinati was optimistic that AI could liberate human employees to work on more productive things. “Their skills are now being uplifted to do other things, whether it’s supervisory work or … learning other skills which allow them to support higher order tasks in the development cycle,” he said.