Despite anxieties about AI replacing human workers, a new study from MIT suggests widespread job losses due to automation are unlikely soon. Researchers found that while AI can potentially handle many tasks, the cost of replacing humans with machines in most cases isn’t currently economically viable for employers.
Only about 23% of wages paid for AI-automatable jobs make economic sense to replace with machines right now, the study revealed. This indicates a slower pace of job disruption compared to earlier predictions, offering policymakers more time to prepare for potential impacts.
“Humans are often the more cost-effective option for now,” says study author Neil Thompson. “While AI has potential, headlines about robots taking jobs should consider the economic realities of implementing these systems.”
The study analyzed jobs susceptible to AI, particularly in computer vision tasks. Researchers compared human wages to potential automation costs, finding that tasks like retail inventory checks are still cheaper for humans.
This gradual pace of AI integration mirrors past technological disruptions like the rise of manufacturing, allowing for proactive adaptation by policymakers, businesses, and workers. While the IMF recently warned of 40% of jobs being affected by AI globally, the new research provides a more nuanced timeline for potential impacts.
“This study allows us to quantify the pace of worker displacement,” says Thompson, “helping develop concrete retraining plans to address the future challenges of AI.”
The findings offer reassurance amidst concerns about AI automation, suggesting a slower transition and more time for preparation and adaptation in the face of technological advancements.
Source: PhilNews24 | January 25, 2024
