ChatGPT, which was developed by OpenAI, was launched in November 2022. Within two months, it had more than 100 million users.
In this article, we shall look at who is leveraging this *generative AI tool, which workers use it to help in their work, which don’t, and why.
* Generative AI is artificial intelligence that creates new content, such as text, images, or code, by learning from existing data and generating outputs similar to human-created work. Examples include ChatGPT, DALL-E, Midjourney, and GitHub Copilot.
Some Embrace Generative AI, Others Don’t
In an article published in the Chicago Booth Review, which belongs to The University of Chicago, Rebecca Stropoli wrote the following about workers using AI:
“A subset said the technology could help them work faster—but many still won’t use it.”
Anders Humlum, from Chicago Booth, and Emilie Vestergaard, a PhD student from the University of Copenhagen, say that younger, less experienced, higher achieving employees are using the AI platform developed by OpenAI.
The researchers carried out a large-scale survey in Denmark and found that half of workers in fields where AI is significantly changing or automating tasks have tried it.
They also found that female workers were twenty percentage points less likely than their male counterparts to use the tool. They described this difference as a “staggering gender gap.”
The researchers say that the results suggest that employers who hope to reap the full benefits from the new AI tool could more actively encourage their employees to use it.
Other Generative AI Tools
In their study, Vestergaard and Humlum focused on ChatGPT as “the icon of Generative AI.” However, most major tech companies have developed or are developing their own AI-powered platforms and tools. Examples include:
- Gemini by Google: On the market.
- Copilot by Microsoft: On the market.
- Llama by Meta: On the market.
- Claude by Anthropic: On the market.
- Apple Intelligence by Apple: On the market.
- Ernie Bot by Baidu: On the market.
- Tongyi Qianwen by Alibaba: On the market.
- Mistral by Mistral AI: On the market.
- Grok by xAI: On the market.
The Survey
The survey, which involved 18,000 workers between November 2023 and January 2024, was a collaboration with the Danish government’s department of statistics.
The researchers focused on eleven occupations that are highly exposed to AI: customer support staff, financial advisers, teachers, software developers, accountants, office clerks, marketers, IT workers, human resource professionals, journalists, and legal professionals. All of them were white-collar occupations.
They linked the survey results to official records about the respondents, including their earnings, education, job histories, and wealth.
Seventy-nine percent of software developers said they use ChatGPT in their work, compared to just 34% of financial advisers. Most respondents said they were aware of ChatGPT. The survey also found disparities by achievement level, age, and gender.
Most Likely ChatGPT Users
The researchers found that younger, less experienced employees were most likely to use the generative AI platform.
For each additional year in a respondent’s age, the likelihood of using AI decreased by 1 percentage point, while each year of experience reduced the likelihood by 0.7 percentage points.
These findings aligned with previous studies that showed that employees with more to learn on the job also had more to gain from ChatGPT or other generative AI tools.
Despite their age and relative inexperience in the workforce, the employees who were most likely to use ChatGPT were also among the higher earners, even before the technology became available.
This suggests that higher-achieving individuals may be more likely to use generative AI, especially those with more education and higher academic performance.
The Generative AI Gender Gap
The gender gap results came from a comparison of males and females in the same professions, workplaces, and a similar mix of tasks.
Women who said there was a barrier to adopting ChatGPT were most likely to say so because they had not had adequate training.
Men, on the other hand, were more likely to cite employer restrictions or concerns about data confidentiality as reasons for not using ChatGPT.
Employers need to do more than simply offer introductory materials to all their workers, the researchers explained.
-
Requesting Training
Even though women said they needed more training, they were less likely to sign up to receive information on ChatGPT usage in their jobs than their male colleagues.
Mixed Time-Saving Potential
Respondents, on average, estimated that generative AI cut their work time by 50% for more than one-third of their tasks.
However, they were twice as likely to report that ChatGPT offered less significant time savings for tasks performed by higher-skilled workers.
Surprisingly, 50% to 60% of those who thought that ChatGPT could cut their time to perform a task by 50% said they had no intention of using it.
When Vestergaard and Humlum told some workers directly how much time ChatGPT could save them, according to experts, they became more positive about time savings.
However, a follow-up survey showed that their likelihood of adopting the technology had not increased.
Reticence not Due to Fear of Being Replaced
The disconnect between employees’ beliefs in ChatGPT’s benefit and their likelihood of adopting it is not driven by “existential fears” of losing their jobs to generative AI, according to the research findings.
-
Employer Policies
Only one-in-ten workers said that being replaced by AI was a concern. The survey findings suggest that employer policies are creating barriers. Forty percent of employees wanted more training while thirty-five percent said that their employers actively restricted usage.
Employers or government programs should provide specific guidelines for workers, as well as targeted training on the technology, the researchers suggest.
Workers who earn less and have lower levels of education, even if they have more job experience, may need extra support to learn ChatGPT. Women may also require focused efforts to help close the gender gap.
-
Maybe Loosen Policies Regarding Generative AI
The researchers suggest that employers that restrict the use of ChatGPT or other generative AI tools consider loosening their policies. They could also re-organize workflows around the tools, they add.
Workers who understand the potential gains in efficiency and productivity from generative AI may be more inclined to use it. This could help boost company profits and growth of the economy in general.