Maastricht Journal of Politics & Economics
AI & The Labour Market: A Workforce Revolution on our Doorstep?
Updated: Jan 23
By Siavash Mohades
The introduction of ChatGPT is a glimpse of the AI explosion coming to all businesses and individuals in every industry or country. Not keeping up with it and not educating labour on how to remain competitive forces both businesses and countries into high-level costs and even bankruptcy. Market experts expect a complex transition process with unknown outcomes in the medium term for the firms and labour, possibly until the human resources become efficient AI users and developers rather than manual workforce rivals.
The Outrage: Concerns and Appreciations
You may have noticed the outrage of ChatGPT on the internet: from simple letters and emails to complicated programming tasks and data analytics, ChatGPT has shown significant potential to replicate human labour, if not improve upon it, with far lower costs. This platform – recently launched by OpenAI – is a large language model trained on a massive trove of online information to create responses to various questions and tasks in any field. In other words, it can use an algorithm on its extensive database of humane information and generate reasonable, decently accurate responses.
From academia to the writing industry to human resources management, there are lights and sparkles of both hope and dark spots of despair. Similar to any other technological outbreak, ChatGPT has raised concerns about the future of those whom it might replace. It can learn – even personalised – patterns of writing, socialising and human interaction to write your daily emails to your colleagues and clients, university essays or the Python codes of your data analysis.
With its potential partial acquisition by Microsoft, it is expected to be soon incorporated into Microsoft Office 365 and benefit from more powerful servers that prevent it from crashing – as was the case in the last week – and also faster than it currently operates. Whether it is the ultimate product of AI for daily work and life or only a preliminary baby step of AI incorporation in our lives, we will soon be using their services just like we did with Google Search Engine.
Academics and Businesses: Opportunities and Threats
Academic users were the first to raise concerns regarding plagiarism and students using ChatGPT to write essays and answer take-home exams. The concerns continued to rise as users tried out different tasks and requests after the OpenAI in different fields and levels of studies. Users quickly figured out how it is capable of generating programs that analyse and train your data for a specific purpose. More strangely, it can write unique essays resembling an analytic perspective and correct writing style - even though it still lacks some humane nature of writing.
It can solve math, it understands Latex input for equation systems and its guesses of more complex mathematical problems, for instance, stochastic integrals or differential equations, have constantly been improving over time. Its versatile "skills" could change the way students are evaluated once and for all, and the educational systems might need a fundamental step forward to find ways to incorporate AI in learning and avoid its inevitable harm to students' learning process.
The business sector will see AI quite differently. AI is no different from a tractor on a farm: it automates the production process. For instance, the medical and manufacturing sectors have benefitted the most in recent years from AI-based automation. Their generic labour share has decreased significantly while their tech-expert labour share has increased by many folds. Thus, their respective products have become more capital-intensive, and their use of high-tech facilities skyrocketed.
Complex products are produced cheaper, become more accessible for more customers, and production externalities have diminished. In principle, these companies were actively reducing their costs by different means. Labour is quite costly for the firms – especially with welfare obligations that, in some regions, companies must bear. Automation reduces the demand for lower-productive labour.
Politics of AI: Will Unemployed Voters Rage?
AI shows up in the political discussion, similar to how tractors were a hot political topic almost a century ago. Policymakers have often felt threatened by country-wide automation processes. In the short term, large groups of individuals lose their jobs as keeping them is too costly for companies.
Take the case of a paper company secretary: their job was initially threatened by automatic phone-call directors, and now AI seeks to reply the emails instead of them – almost for free! Such layoffs in the short term may be perceived as signals that the economy cannot create jobs for everyone, and the political costs could be extensive for the incumbent governments.
When facing such challenges imposed by automation, many countries proposed labour protection measures to make the firing process difficult, costly and by some means "caring" from the employees' point of view. Companies will take action if only the costs of holding on to the labour exceed the benefits of adapting to the technology. Such decisions are not unique to the automation threats: even an increase in obligations regarding labour's welfare might lead to massive layoffs, as has been the case in big businesses in the U.S. since December.
All in all, ChatGPT and AI's role in replicating human action and behaviour is there to stay. Neglecting it would make firms and industries bear high opportunity costs and less market power through higher offering prices and marginal costs – which more technically construct each firm's markup. Possible challenges for governments would probably be to remain longsighted and invest in education, mainly primary and secondary schools.
By doing so, those who cannot or do not want to attend college and higher education will remain competitive in the labour market based on the skills they learn in school. This could be done for example by encouraging programming classes in the most popular languages in schools, for students to be driven into this massively growing part of the labour market.
AI is growing rapidly - more than ever -, and this is why new ways governments and policy institutions need to learn and develop new ways of incorporating AI into the education system. This goes hand in hand with lowering labour market barriers and allowing the hidden hand mechanisms of markets in determining labour demand and wages, precisely like they did when tractors replaced human manual labour on farms.
Sources: CNN, Forbes, OpenAI, Edutopia, Mindmatters & Youtube (The Office)
Written by Siavash Mohades