Effectively integrating human work with artificial intelligence

4 March 2024

The recent emergence of programmes such as ChatGPT has caused the resurgence of doomsday scenarios whereby artificial intelligence (AI) conquers the world, subordinating humankind to the dictates of their autogenerated algorithms.

However, there are other scenarios that paint a rosier picture of people being freed from the drudgery of mundane work to enjoy more leisure time. Ultimately, neither of these scenarios are likely. Yet, the degree to which technological freedom or subjugation will come to pass depends more on the actions of humans than anything done by a machine.

In my new book, The Future of Work in Diverse Economic Systems: The Varieties of Capitalism Perspective, I argue that AI is only capable of performing standardised tasks, albeit at a faster and more precise rate than humans. It is how humans will interact with AI, rather than AI per se, that will determine our future.

What can AI do?

To understand how humans can shape the impact of AI on our lives, we first need to be clear about what AI is really capable of doing.

Even when AI systems seem to complete tasks autonomously, they are really just applying the rules provided to them by their human programmers. AI is essentially just classifying data and making predictions based on the instructions received from humans.

Although computer programmes can determine patterns, they cannot distinguish between cause and effect. They also cannot adapt to novel situations, because they lack cunning, intuition and the ability to improvise. Even the most sophisticated AI programs, namely neural networks, have capabilities that are more similar to the intelligence of a worm than a human being.

Can we work with AI

With the emergence of AI, the tasks left to humans are those that involve the analytical skills needed to solve novel problems and adapt to new situations.

A prime example of this synergy can be observed in law offices across the world, enabling them to focus more on the analytical side of their work. AI is simply incapable of performing analytical tasks in the vast majority of jobs.

…or will AI replace us?

Statistics that estimate that over 50% of all jobs in the Organisation for Economic Co-operation and Development could be eliminated by AI overexaggerate its impact. This is because such statistics assume that any job, in which some of its tasks can be eliminated, will be wholly threatened by AI. When we use statistics that define jobs ‘likely to be eliminated by AI’ as those for which 70% of tasks can be automated, only 9% of the jobs in the United States are considered at risk.

Yet, even these studies overlook that the implementation of new technology also leads to the creation of new jobs. For example, the introduction of the automated teller machine actually caused employment in the banking industry to increase, as banks were able to open more branches and offer more services – run by more staff.

Although cost savings from implementing AI can be astronomical, the investments required to buy and install them can be prohibitively expensive. Companies that use AI only to optimise processes, and thereby reduce costs, may not save enough to justify such financial outlays.

Can AI help drive innovation?

AI alone will not help firms innovate. They need workers with analytical skills to undertake these tasks. For example, in the wooden furniture industry, some firms innovate tables simply by developing new legs or other parts. The rest of the components do not have to change, and they can be made using AI. Workers and these new technologies can coexist in such firms and actually complement each other. However, it requires the vast majority of companies to be more innovative and an educational system that can provide workers with these skills. Workers in such factories need to know much more than simply how to push buttons.

Societies would clearly benefit from firms finding ways in which workers and AI can effectively complement each other. If the majority of firms in a country eventually choose to make standardised products based extensively on AI, we may indeed become servants to the machine. However, if companies and societies take the other route, work life can potentially become more interesting for a much wider variety of people.

Share this: