MIT Study Reveals Surprising Findings on AI’s Impact on Jobs
A new research study from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) has unveiled some unexpected findings regarding the impact of AI on jobs. The study aimed to address three key questions: Will AI automate human jobs? If so, which jobs? And when?
Numerous predictions have been made about how AI technologies, particularly large language models, will affect people’s livelihoods and economies in the future. Goldman Sachs estimates that AI could automate 25% of the labor market within a few years, while McKinsey predicts that nearly half of all work will be AI-driven by 2055. Additionally, a survey conducted by the University of Pennsylvania, NYU, and Princeton suggests that ChatGPT alone could impact approximately 80% of jobs.
Contrary to expectations, the MIT researchers discovered that the majority of jobs previously identified as being at risk of AI displacement are not currently economically beneficial to automate. The study challenges the notion that AI will rapidly and dramatically replace human workers.
According to Neil Thompson, a research scientist at MIT CSAIL and a co-author of the study, while there is a significant potential for AI to automate tasks, many of these tasks are not yet attractive for automation. The researchers focused specifically on jobs requiring visual analysis, such as inspecting product quality on a manufacturing line. They did not investigate the potential impact of text- and image-generating models on the economy.
The researchers surveyed workers to understand the specific tasks an AI system would need to accomplish to fully replace their jobs. They then estimated the cost of building such a system and assessed whether businesses would be willing to invest in it. The study provides an example of a baker whose job involves spending around 6% of their time checking food quality, a task that could be automated. However, the researchers estimated that the cost of implementing the necessary AI system would outweigh the potential savings.
The study does have limitations, which the researchers acknowledge. It does not consider cases where AI can augment human labor or create new tasks and jobs. Additionally, it does not factor in the cost savings that can come from pre-trained models. However, the researchers insist that their conclusions are not influenced by the study’s sponsor, the MIT-IBM Watson AI Lab.
The study’s findings suggest that AI job automation will occur gradually over time rather than in an abrupt and widespread manner. It emphasizes the importance of preparing for AI job automation through policy initiatives while also highlighting the need to decrease the costs of AI deployment and increase its scope for economic attractiveness.
In conclusion, the MIT study provides a nuanced perspective on the impact of AI on jobs. While the potential for automation exists, the study suggests that the process will unfold slowly and that there is time for policy initiatives to address the challenges posed by AI job automation.
