When the steam engine transformed industry in the 19^th century, economists redefined how they measured growth and economic productivity. Today, two centuries later, artificial intelligence poses a similar challenge, except now the developments and changes come from AI algorithms and not machinery or infrastructure. The world of AI automation is growing rapidly and is enabling people to fully automate tasks that were once performed by humans. Some are eager to welcome this advance in technology and think it will encourage a new productivity boom, while others are sceptical of becoming over-reliant on AI and losing the human aspect of the workforce. Artificial intelligence can create work forces that never sleep, never get sick and never miss deadlines; as AI begins to function less like a tool and more like a worker, economists are urged to rethink what productivity means in the age of AI-powered digital workers.  

The ‘productivity puzzle’ and the rise of AI: 

The ‘productivity puzzle’ refers to the slow growth in output per worker despite rapid advances in technology. In developed economies, labour productivity growth has averaged around 1.5% annually since 2005 (down from the 3% rates seen in the 1990s and early 2000s). This slow growth is potentially detrimental because ultimately productivity is the main driver of rising living standards and GDP per capita in any economy, including in the UK. This lower growth rate eventually leads to reductions in living standards, slower GDP growth and fiscal pressure for the government. As a result, people fear that AI will follow the same slow growth in productivity that was seen in the computer age. However, many believe that AI is different due to its agentive and generative nature. This means it is able to learn from mistakes, develop and adapt in order to better solve your problems, and perform a vast number of tasks. AI can be a receptionist at the front desk, a member of the sales team on the phone, or even your full-time personal assistant. Adoption of artificial intelligence in the workplace has accelerated at an unprecedented rate, with many offices worldwide using large language models (LLMs) such as ChatGPT to aid them in daily tasks. If AI continues to progress and automate human tasks at a wider level, economists may have to consider AI as a new factor to consider when calculating productivity alongside labour, capital, land and enterprise. 

AI’s productivity potential: 

It is clear that AI could offer a much-needed boost to productivity, both globally and within the UK. Studies show that users of generative AI are able to reduce their weekly work hours by as much as 5.4%. This is equivalent to over two hours per week saved for a full-time employee. This is due to the automation of mundane and time-consuming tasks, which allows workers to focus on the more human-driven aspects of their work and also have more time in their lives. Therefore, many people believe that harnessing the power of AI will increase innovation, product development and entrepreneurship in economies around the world. Goldman Sachs projects that, if AI is fully integrated, it could increase labour productivity by up to 15%. Statistics like this could greatly decrease the long-standing productivity gap seen in the UK when compared with countries such as Germany and France. However, the main challenge is the integration of AI across the whole economy, instead of just in tech-heavy industries such as finance and professional IT services in which the UK is already globally competitive. 

Limits, risks and the measurement gap: 

Despite the clear potential of AI, its positive impact on productivity is not a guarantee. It is important to remember that generative AI is still very new and therefore many people are still learning how to use it and implement it properly and effectively. There are currently many examples of low-quality AI integration into businesses in the UK. Unfortunately, many of the UK’s small and medium-sized enterprises lack the capital and/or expertise to implement these new tools effectively. This lack of expertise is highlighted by how 74% of small to medium-sized business employees report using AI tools, while only a third of employees report receiving formal instruction on how to use them. Essential skills such as prompt engineering, LLM use and Application Programming Interfaces (APIs) are often neglected, resulting in people using these tools wrong and failing to get the most out of them. AI tools are becoming increasingly user-friendly and easy to use (even for people with no coding experience), but they still require people to have a base knowledge of artificial intelligence to get the most out of it. Conversely, the growing accessibility of AI could also encourage unauthorised use by workers, which could lead to threats to the overall security of companies and data. Economists must consider whether AI literacy/education is accessible enough and whether the improvements it provides in the workplace would translate into measurable growth in the UK’s official productivity figures.  

Rethinking productivity – AI as a digital labour force: 

If AI is treated less like a tool and instead as a form of ‘digital labour’, then economists will need to rethink how productivity is defined and measured; traditional metrics such as output per worker and output per hour worked may no longer be clear measures of productivity in an age where algorithms can produce outputs too (often more efficiently and accurately). The question that arises is whether AI should be counted as an extension of human labour, or a distinct factor in productivity. Despite possible benefits to productivity in the future, AI could replace thousands of workers which would eventually lead to the overall national income to decline further, worsening the wealth gap. This is because many of the gains from AI will flow to those who own or are able to develop the technology, while many less-skilled workers lose their job and income, further increasing the wealth gap often seen in developed economies. Ultimately, by acting as a new labour force, AI has the potential to reshape economies worldwide through the increase in productivity that it may be able to create and the subsequent rise in living standards which would be seen as a result of its ability to increase overall national productivity. 

Conclusion: 

Many people believe that AI will simply ‘take their jobs’. Some believe that AI could become a replacement for the modern idea of a worker; however, others believe that AI can be harnessed as a tool that can greatly increase productivity for economies across the world. Overall, the rise of AI automation will lead to greater labour efficiency, more innovation and greater economic output in the future, ultimately benefiting millions of people and raising productivity worldwide.