Initial productivity gains from AI may prove transient, new study reveals.
<div class="media-with-label__label">
Credit: Arts Illustrated Studios / Shutterstock </div>
</figure>
</div>
</div>
</div>
</div>Widespread assertions suggest AI will streamline tasks and dramatically elevate productivity. For instance, PwC projects that AI adoption could increase global economic output by an additional 15% within the coming decade.
However, the reality might not be as optimistic as proponents of artificial intelligence often portray. Recent findings from Harvard Business Review (HBR) indicate that the perceived advantages of AI might have been overstated.
The study observed that staff utilizing AI applications tended to complete work more quickly and for extended periods, managing a greater number of responsibilities. “Such shifts are unsustainable, potentially causing an escalation of workload, mental exhaustion, professional burnout, and impaired judgment. The initial boost in productivity can subsequently diminish, leading to a decline in work quality, increased staff turnover, and additional challenges,” noted the researchers, Aruna Ranganathan and Xingqi Maggie Ye.
These discoveries align with previous expectations. A study conducted last year by researchers at the University of Chicago and the University of Copenhagen revealed that AI chatbots provided only about an hour of weekly savings for employees, often generating new duties that effectively negated any gained time.
The HBR study identified three primary mechanisms through which AI has heightened work demands: an increase in task scope, the erosion of distinctions between professional and personal life, and a rise in concurrent task management.
This serves as a further indication that although AI simplifies mundane operations, its adoption introduces various complications. “The critical query for organizations is not if AI will transform work, but rather if they will proactively guide that transformation — or passively allow it to mold their operations,” the researchers conclude.
Originally published by CIO.
