As ChatGPT and other generative AI tools have taken the world by storm over the past 18 months or so, there has been an understandable desire to examine how these tools are being used, how effective they’ve been, and who is getting the most from them. To date, many studies have suggested that the tools “equalize” things, with lower-skilled workers benefitting more than higher-skilled workers, thus potentially raising their productivity.
A recent study from Chicago Booth explores whether this heuristic results in more usage from lower-skilled workers. The study, which surveyed around 100,000 workers, examined how ChatGPT was being used by workers in Denmark. The research found there was widespread adoption of the technology, with nearly 80% of software developers using it in some way. Across the professions, around half of workers said that they’d used it at some point, with around a third currently using it.
Interestingly, despite this strong usage, very few respondents said they were using the subscription version of the software. Indeed, just 6% said they were paying to use ChatGPT.
The Matthew Effect
It’s also of note that far from acting as a way of “leveling up” the workplace, it was actually more likely to emphasize the divide. The study found that, whereas 78% of higher-skilled workers adopted the technology, this percentage fell to just 36% among lower-skilled workers.
This kind of “Matthew Effect” has been seen in various technologies in recent years, with higher-skilled individuals better able to capitalize on the potential and therefore extending the gap between them and lower-skilled workers, even when the technology had minimal cost barriers in the way of more widespread adoption.
The study also found that women were around 20% less likely to use ChatGPT than men were, even when taking into account things like experience and the roles performed.
Productivity gains
The results do suggest, however, that there are some productivity gains to be had from using ChatGPT. The survey found that workers estimate that ChatGPT can cut working time in about a third of their tasks. They believe it offers less time savings for workers with more expertise.
This contrasts with the findings that higher-skilled workers are using it more often, however, although it should be said that the productivity gains are largely perceived rather than anything officially recorded.
This is perhaps explained by the fact that while many suggested that ChatGPT helped them to do things faster, they weren’t really taking on additional work to fill their free time.
A role to play
What might the findings tell us about the role generative AI can play in the workplace? They suggest that while these tools could potentially improve productivity and efficiency, the gains from them are unlikely to be distributed evenly across the workforce.
Instead, higher-skilled workers are more likely to adopt and benefit from these technologies, which could potentially widen any existing inequalities both within the workplace and the economy more broadly. What’s more, the gender gap in the technology’s adoption points to broader issues of accessibility and inclusion that need to be addressed.
To overcome these challenges, the authors believe that companies might consider providing training and other kinds of support to help lower-skilled workers and other underrepresented groups make effective use of these new technologies.
By providing more equitable access to technology and ensuring that all workers have the necessary skills to leverage these tools, it may be possible to create a more balanced playing field where the gains are spread more evenly.
Uneven gains
What’s more, a simple understanding that the perceived productivity gains from AI tools like ChatGPT can vary depending on the skill level and task complexity can be crucial for managers at all levels of the organization. Businesses and policymakers should consider these nuances when designing interventions and support systems to maximize the benefits of AI technology across all levels of the workforce.
The study reminds us that while there has been considerable interest in ChatGPT, and other generative AI tools, the gains being made from it are perhaps not as “equalizing” as many have predicted. Higher-skilled workers are more likely to benefit from these technologies, and there are notable gender disparities in adoption.
Addressing these issues through targeted support and training can help ensure that the benefits of AI are more widely shared, which is likely to ultimately contribute to a more equitable and productive workforce at the same time.
