2023 was the year AI really made its mark. It’s been a wild ride for everyone from office workers to big companies, with even the experts scratching their heads. Imagine computers that can whip up a business report, craft a speech, or write a smooth email in a flash. It’s impressive, sure, but it’s got a lot of us asking, “Is my job safe?” Predicting the future of jobs is tricky, and guesses often miss the mark. Yet, we don’t need a crystal ball to see how AI, with all its fancy new tools like ChatGPT and Dall-E, is starting to shake things up in the job world. 2023 wasn’t just another year; it was a turning point, giving us a peek at how AI could reshape our work lives.
Last year, three insightful studies shed light on AI’s evolving role in the professional lives of knowledge workers, presenting a mixed bag of outcomes. On one hand, we can see a clear boost in productivity and a promising step toward levelling the playing field in terms of work equality. But on the other, we’re seeing increased competition and a potential dip in earnings.
The first study from the US casts a warning shadow by showing that freelancers on digital platforms are finding it harder to secure jobs and maintain their income levels following the introduction of advanced AI tools like ChatGPT and Dall-E. Within just six months of these AI advancements entering the market, creative professionals in writing and design saw a noticeable downturn in work and wages. While the study stops short of directly linking developments to AI, the timing is hard to ignore.
However, it’s not all negative. A groundbreaking experiment by MIT researchers highlighted AI’s potential to be a boon for the workforce, significantly enhancing efficiency and reducing disparities in work quality. Particularly striking was the finding that workers who at the start of the study were lagging in productivity saw the most notable improvements, suggesting AI might help in redistributing productivity gains more evenly across the workforce.
Adding another layer to the conversation, a detailed investigation with the Boston Consulting Group (BCG) looked into how AI, especially Large Language Models (LLMs), impacts the complex task management of consultants. The results were encouraging: consultants using AI were more productive, completing tasks quicker and with higher quality, across various skill levels. However, the benefits of AI weren’t universal; some complex tasks remained out of AI’s grasp, and not all consultants found AI’s guidance to be beneficial.
Left: Monthly employment for freelancers declines after the introduction of ChatGPT. Source: Hui, X., Reshef, O., & Zhou, L. (2023) Right: Professionals improve the quality of their work when using ChatGPT. Source: Noy & Zhang (2023)
Together, these studies paint a picture of AI’s dual impact in the workplace: it’s a powerful tool that can drive efficiency and even-out productivity levels, but it also introduces new challenges in terms of job competition and wage stability. From what we’ve seen so far, the big question that emerges is: Will AI enhance our work or start replacing it?
To really grapple with the debate between “AI boosting our work versus replacing us”, we need to understand how any new technology reshapes our jobs. History shows us that tech steps in when it can do a job more skillfully and cost-effectively than humans. But that doesn’t mean it wipes out human roles completely. In our daily grind of managing emails, reports, meetings, and negotiations with colleagues, AI might excel in some areas and fall short in others. The more varied and complex our job tasks are, the less likely it is that AI will take over completely.
On this, the BCG study introduced an interesting idea they termed the “jagged technological frontier,” showing that AI’s helpfulness really depends on the task at hand. They talk about “Centaurs” and “Cyborgs” as examples of people working with AI. Centaurs collaborate with AI, splitting tasks to play to each other’s strengths, whereas Cyborgs weave AI more tightly into their day-to-day tasks, constantly interacting with it. These distinct profiles underline a key point: those whose jobs focus on a narrow set of skills, like many online freelancers, might be at greater risk of being replaced by AI. When a job is built around one specific skill and AI learns to do it, the need for a human to do that specific job might decrease. In contrast, AI might become your most powerful colleague, if it only substitutes a limited set of your skills at work.
The idea of working alongside AI—like the Centaur who teams up with AI to improve their overall productivity—shows us how we can not only adapt to changes but also thrive with AI by our side. Learning to work with AI, however, is a skill on its own. Mastering tools like LLMs for writing reports or image generation for graphic design isn’t something you do overnight. Yet, the job market is already on the lookout for people skilled in AI, signalling a shift from a general fear of job loss to a seeking out of AI-savvy professionals, as the latest findings from our research group at the Oxford Internet Institute shows.
One of our recent studies reveals that AI-related skills can significantly boost incomes, with some workers seeing up to a 40% increase in wages. The surge in demand for these skills means that being proficient in AI can be as valuable as holding a university degree, with some job sectors valuing AI expertise even more than traditional academic credentials, as our analysis of one million job offers from the UK indicates. So, what’s the takeaway from the growing impact of AI and the importance of AI skills in the workplace?
Left: The additional wage offered for AI skills is on par with having a PhD. Source: Gonzalez Ehlinger & Stephany (2023) Right: Some in-demand AI skills can significantly increase wages. Source: Stephany & Teutloff (2024) (NLP = Natural Language Processing)
As the workplace continues to evolve, staying ahead means embracing skills that work hand in hand with new technologies. This might mean looking beyond conventional education paths and considering apprenticeships, online courses, or on-the-job training to fill any skill gaps. Employers, too, should lean into skills-based hiring and foster environments that encourage continuous learning. For businesses lacking the means to provide extensive reskilling internally, government support could play a crucial role, perhaps by setting up centres where people can get formal recognition for their self-taught skills, ensuring everyone can navigate the changing landscape of work with confidence.
Watch Dr Fabian Stephany sharing his latest insights on the impact of AI on the future of work.
This project explores how the analysis of online generated labour market data can help us understand the relevance of new skills and sustainability of novel occupations.
This programme supports research in the sphere of AI & Work.
