In recent years, the landscape of technology has evolved at an unprecedented pace, with companies such as OpenAI and Google unveiling advanced Artificial Intelligence (AI) systems that have transformed what was once science fiction into an everyday reality. This rapid change has led to crucial questions surrounding the future of humanity and the workforce. Stanford University-based technology-focused economist Erik Brynjolfsson has been studying the economic impact of these AI systems and their implications on the workforce.
Brynjolfsson, along with MIT economists Danielle Li and Lindsey R. Raymond, recently conducted the first empirical study of the real-world economic effects of AI systems. The researchers examined the impact of ChatGPT, an interactive AI chatbot, on a company’s operations and its workers. They discovered that AI can significantly improve worker productivity and customer satisfaction, while also raising concerns about the potential disruption and implications for income inequality in the future.
The study began when a Fortune 500 company implemented an earlier version of OpenAI’s ChatGPT to assist its customer support agents in resolving technical issues for clients. These agents, based primarily in the Philippines, communicated with customers through online chat windows. With the AI system in place, the agents could consult a separate chat window with the chatbot to receive real-time advice on responding to customers and finding solutions to their problems.
Upon analyzing the performance of over 5,000 agents, the researchers found that AI-enhanced customer service representatives were, on average, 14% more productive, resolving more customer issues per hour. Turnover rates decreased, especially among new hires, and customer satisfaction ratings improved. Interestingly, the study also revealed that it was the less experienced, lower-skilled representatives who saw the most significant gains in job performance, as opposed to the more experienced, highly skilled agents.
Brynjolfsson suggests that this outcome makes sense considering how AI systems work. By learning from highly-rated interactions between agents and customers, the chatbot effectively mimics top performers’ strategies, enabling lower-skilled workers to improve their performance. The question then arises whether the company should be compensating top performers more, as their expertise indirectly benefits all of the company’s customers.
While the study points to the potential for AI to boost the economy’s productivity and improve living standards, Brynjolfsson cautions that there are no guarantees about how the resulting wealth will be distributed. Ensuring that gains are shared fairly may require changes in tax policies or company strategies. Moreover, AI’s impact on the labor market may differ from previous waves of technological change, potentially benefiting lower-skilled workers who were left behind in the past.
On one hand, AI could reduce inequality by improving the skills of lower performers. On the other hand, it could decrease the value of specialized skills and experience, making it easier for anyone to perform certain jobs and thus lowering their wage premiums. Alternatively, AI could exacerbate inequality by further enriching the companies that own these systems or by enabling businesses to replace more workers with intelligent machines. The full implications of AI on the workforce remain uncertain, but its potential to perform non-routine tasks and adapt on the fly signals a distinct departure from previous technologies.
A version of this article was originally posted on NPR: https://www.npr.org/sections/money/2023/05/02/1172791281/this-company-adopted-ai-heres-what-happened-to-its-human-workers
Thoughts from IntelliChat: With an ongoing struggle to find entry-level workers and the need for solutions to empower and train current staff, AI Support and Training provides a solution. In addition, the long training times of new tech support employees can be drastically reduced through the use of AI. Although there may be some currently evident disparities in the usefulness of AI in comparison to an entry-level vs senior tech, as AI improves these disparities will likely decrease. Regardless of the advancements of AI, solutions like IntelliChat will greatly alleviate current hiring and retention pressures for companies with medium to high-tech support needs.
