Recent research conducted by Boston Consulting Group (BCG) and Wharton Business School has elucidated the potential productivity boost that can be achieved through the integration of artificial intelligence (AI) in the workplace, particularly through the use of chatbots such as ChatGPT.
BCG consultants who utilised ChatGPT for a variety of tasks demonstrated a notable increase in productivity and quality of outputs in contrast to their non-AI using counterparts. The AI enabled consultants were found to finish 12.2% more tasks, complete tasks 25.1% more quickly and produce 40% higher quality results. The study involved a range of tasks common to the consulting field, including writing, marketing, persuasion and creativity tasks.
Leveling the Playing Field with AI
It was also found that AI acts as a skill-leveler, particularly benefiting the weakest consultants, who displayed a remarkable 43% performance improvement when using AI. This aspect, coupled with the findings of a similar study from Stanford and MIT, suggests that AI could elevate lower-performing employees closer to the level of their higher-performing colleagues.
Distribution of output quality across tasks with and without using AI assistance (Source: Ethan Mollick)
“AI at Work”, a report from job placement firm Indeed, further highlighted AI's influence in the job market. An analysis of more than 55 million job postings and 2,600 job skills revealed that around 20% of jobs are highly exposed to the impact of AI – meaning AI could perform 80% or more of the skills required for the position. Moreover, 45% of job listings are moderately exposed, with AI capable of executing between 50-80% of the skills needed.
However, a pressing question remains: Can AI handle complex tasks that traditionally require human intuition and creativity? An experiment by Section School seemingly supports this, with the Claude AI chatbot demonstrating an ability to provide feedback on a board slide deck – a complex and nuanced task traditionally handled by human board members.
Largely, the implications of these findings suggest that AI can boost business productivity and revenues, with the possibility for companies to achieve more with an equal number of employees or even potentially downsizing in certain areas. Highly skilled employees could allocate their time to tasks that AI cannot perform, broadening the range of skills in the workforce.
The Double-Edged Sword of AI Implementation
A separate paper by Fabrizio Dell'Acqua from Harvard suggests that an overreliance on AI could lead to individuals becoming less skilled and observant in their own roles – a potential drawback of AI implementation.
These findings highlight that while AI can dramatically enhance productivity and task quality, it must be judiciously implemented, so as not to deteriorate human capabilities. Striking the right balance between human and AI involvement at work is critical to ensure optimal engagement and productivity.