Generative AI Helps Workers Learn New Skills, Says Study by BCG and Boston University

A recent scientific study revealed that Generative AI (GenAI) can significantly expand the capabilities of workers, enabling them to perform tasks beyond their current skill sets. The study was conducted by the BCG Henderson Institute, in collaboration with BCG X and Emma Wiles from Boston University.

Study Details

The study, titled “GenAI Doesn’t Just Increase Productivity. It Expands Capabilities”, explores how GenAI can enhance workers’ ability to take on new tasks, challenging the traditional boundaries of knowledge work. The findings highlighted the potential of GenAI-augmented workers to tackle new tasks, even when they had little to no prior experience in coding or statistics.

Experiment and Results

The experiment involved 480 BCG consultants, who completed two of three tasks typically performed by data scientists: writing Python code to merge datasets, building predictive models, and validating ChatGPT-generated statistical analyses. These tasks were designed to be difficult for non-specialists and required manual input from the participants, despite the assistance of the GenAI tool, powered by GPT-4 with advanced data analysis features. Their results were compared with those of 44 BCG data scientists who completed the tasks without AI support.

As per the study, the participants using GenAI demonstrated a notable improvement in coding tasks, scoring an average of 86 per cent of the benchmark set by data scientists, a 49-percentage point improvement over those without AI support. It also revealed that GenAI-assisted participants also completed tasks roughly 10 per cent faster than data scientists.

“When using GenAI, the consultants in the study were able to instantly expand their aptitude for new tasks. Even when they had no experience in coding or statistics, consultants with access to GenAI were able to write code, appropriately apply machine learning models, and correct erroneous statistical processes,” said the study.

Challenges and Insights

However, the tool was less effective in the predictive analytics task, where both consultants and the GenAI tool struggled to achieve optimal results. Despite this, GenAI acted as a powerful brainstorming partner, allowing participants to combine their knowledge with AI to discover new techniques and improve problem-solving.

“For the predictive analytics task, the experiment participants encountered a significant challenge: neither they nor the GenAI tool had a high level of proficiency in this area,” the study noted.

One of the key takeaways of the study is that GenAI, while expanding capabilities, does not automatically reskill workers. Participants did not show significant learning gains, indicating that “doing” tasks with AI support does not equate to “learning to do” those tasks independently.

“In an era of unprecedented technological disruption, the global study The Rise of the GenAI-Augmented Worker reveals a pivotal shift in the knowledge workforce. Technology workers are not just learning faster–they are enhancing their skills in ways that significantly boost productivity and innovation,” said Nipun Kalra, Managing Director and Partner at BCG India.