[Photo from Getty Images]
A recent analysis by the Bank of Korea revealed that while the introduction of artificial intelligence (AI) has reduced work hours by an average of 1.5 hours per week, it has not translated into increased productivity. This phenomenon, termed 'productivity disconnect,' arises because AI implementation has not led to organizational redesign or workforce reallocation.
According to the 'BOK Issue Note: Does AI Increase Productivity? An Analysis of Initial Three-Year Effects' released on June 7, AI usage has been shown to shorten work hours by an average of 3.8%. This equates to a savings of approximately 1.5 hours per week.
The time-saving effects of AI are particularly pronounced among low-skilled workers and those who heavily utilize AI. The Bank of Korea noted that as proficiency in AI increases, the marginal efficiency gains from technology adoption also rise. Furthermore, AI helps to mitigate productivity gaps among low-skilled workers by compensating for their lack of experience.
When translating these reductions in work hours into potential productivity gains, the Bank estimates an increase of about 1.0%. However, the time saved through AI usage has not resulted in actual production increases. The correlation between individual reductions in work hours and increases in work output was estimated to be zero.
The Bank explained that while AI has improved efficiency at the individual task level, its benefits have not spread to enhance overall workflow, organizational structure, or workforce reallocation, leading to the observed 'productivity disconnect.' Bottlenecks in production processes and distortions in performance reward systems were also identified as factors hindering productivity transformation.
Conversely, groups with strong performance incentives and high job autonomy have seen productivity gains. This includes self-employed individuals, professionals, and intensive AI users.
By age group, younger workers (ages 15-39) increased their output by approximately 0.6 percentage points more than those aged 50-64. The Bank interprets this as a result of younger workers' greater adaptability to digital technologies, allowing them to connect AI usage more effectively to productive activities.
In terms of occupation, professionals increased their output by 0.7 percentage points more than office workers, while the top 50% of AI users demonstrated a 0.5 percentage point greater improvement in productivity compared to the lower half. This suggests that higher intensity of AI use is likely to overcome initial friction costs, such as learning expenses and validation burdens, leading to tangible productivity enhancements.
Oh Sam-il, head of the Bank's Employment Research Team, stated, "Currently, AI has entered the 'efficiency' stage but has not yet fully transitioned to the 'productivity' stage. This can be viewed as a typical transition process in the early stages of adopting a general-purpose technology. Depending on future policy responses and changes in corporate organization and labor market structures, the productivity trajectory could change significantly."
He added, "To realize the productivity effects of AI, it is crucial to redesign work processes and organizational structures, reallocate job roles, and establish performance-based incentive systems. Continuous monitoring of changes in the skill development pathways for younger workers is also necessary."
According to the 'BOK Issue Note: Does AI Increase Productivity? An Analysis of Initial Three-Year Effects' released on June 7, AI usage has been shown to shorten work hours by an average of 3.8%. This equates to a savings of approximately 1.5 hours per week.
The time-saving effects of AI are particularly pronounced among low-skilled workers and those who heavily utilize AI. The Bank of Korea noted that as proficiency in AI increases, the marginal efficiency gains from technology adoption also rise. Furthermore, AI helps to mitigate productivity gaps among low-skilled workers by compensating for their lack of experience.
When translating these reductions in work hours into potential productivity gains, the Bank estimates an increase of about 1.0%. However, the time saved through AI usage has not resulted in actual production increases. The correlation between individual reductions in work hours and increases in work output was estimated to be zero.
The Bank explained that while AI has improved efficiency at the individual task level, its benefits have not spread to enhance overall workflow, organizational structure, or workforce reallocation, leading to the observed 'productivity disconnect.' Bottlenecks in production processes and distortions in performance reward systems were also identified as factors hindering productivity transformation.
Conversely, groups with strong performance incentives and high job autonomy have seen productivity gains. This includes self-employed individuals, professionals, and intensive AI users.
By age group, younger workers (ages 15-39) increased their output by approximately 0.6 percentage points more than those aged 50-64. The Bank interprets this as a result of younger workers' greater adaptability to digital technologies, allowing them to connect AI usage more effectively to productive activities.
In terms of occupation, professionals increased their output by 0.7 percentage points more than office workers, while the top 50% of AI users demonstrated a 0.5 percentage point greater improvement in productivity compared to the lower half. This suggests that higher intensity of AI use is likely to overcome initial friction costs, such as learning expenses and validation burdens, leading to tangible productivity enhancements.
Oh Sam-il, head of the Bank's Employment Research Team, stated, "Currently, AI has entered the 'efficiency' stage but has not yet fully transitioned to the 'productivity' stage. This can be viewed as a typical transition process in the early stages of adopting a general-purpose technology. Depending on future policy responses and changes in corporate organization and labor market structures, the productivity trajectory could change significantly."
He added, "To realize the productivity effects of AI, it is crucial to redesign work processes and organizational structures, reallocate job roles, and establish performance-based incentive systems. Continuous monitoring of changes in the skill development pathways for younger workers is also necessary."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.