This brief editorial essay was originally planned to be about multidisciplinarity and interdisciplinarity, but AI made me change the topic – sort of! You can already see what is happening here – AI *made me* change my plans! How powerful is that? This kind of imagery of an all-powerful technology changing lives has been rampant. But what is really happening is that technology has been reified – it has been given attributes of capability and decision, agency and freewill. However, such reification is highly problematic.
We are being swamped with claims for AI – in print media, AV media, social media …. It is a rare day when we do not see a headline.
AI is taking our jobs!!
AI will make good businesses great!
AI – the silver bullet for improving productivity.
AI is increasing the surveillance of workers to a dangerous degree.
Perhaps like many of you, I was becoming uncomfortable with all of these claims. As a grubby empiricist, I wanted material evidence. I wanted to know why all these emphatic claims are made with force and certainty, and it was important to understand this for several reasons. Perhaps the most important reason for this essay has been that AI has been sold as a boon - and moreso as a boom, a dominating and unstoppable force. (Of course, booms are not always built on substance (tulips – or dot.com anyone? But that is another story). We are all being bombarded with claims about the potential of AI to revolutionise work, production, and creativity. The rhetoric is so forceful that it is taking on the mantle of reality. AI will do all those things listed above because everyone says so.
No – too many sweeping assertions. I needed to know more about AI.
I started my search with a string of academic articles, and then The Conversation (which publishes articles by experts in terms accessible to the non-experts). These all proved helpful and gave me insights into the development and ownership of AI, but I needed more.
So, of course, like all not-so-good modern researchers, I asked AI!! My trusty computer has an AI generator programme, so I asked – What is AI going to do for workers? What are the benefits and challenges?
And that AI generator programme responded
”AI is not simply a threat or a benefit – it’s a transformative force ....”.
That was even worse than all the simplistic assertions and weak assumptions I had been seeing in the media. Doubts crept in. Perhaps AI was this great force ….??
No – let’s be realistic here.
I am sorry to my AI generator, I am sorry to the sub-editors of news sites (and they are perhaps one and the same in some media?), I am sorry to the breathless commentators and AI boomsayers. But it is not correct. Regardless of all the claims, AI is none of these things.
AI has no free will, no intelligence, and no capacity to take action.
At base, AI is “large scale statistical pattern matching”. Yes, it can get larger and larger as IT and data centre capabilities grow, but really, AI is autocorrect on steroids (and we all know how irritating that is!)
Seriously, let’s cut back on all this reification. AI is not a thing – it cannot spy, it cannot of itself increase productivity or stalk workers to a terrifying degree. Rather, AI is a series of software tools. The changes in work and production processes and procedures can only occur when and if businesses and employers decide to implement AI; they also decide what to implement – how much surveillance or advanced planning is needed.
This is not new. The history of society is the history of employers choosing to replace or control workers with technology. The industrial revolution, the development of steam, electricity, and automation – these were phenomena for employers and businesses to choose. Like AI, neither steam nor electricity nor automation chose to increase speed, productivity or surveillance. Today, it is the same with the expansion of generative AI. Despite claims made for AI’s capacity to build itself, in reality, the extent and use of AI depends on when the (gigantic) IT businesses enable changes to be implemented or institute the generative technical advances to be made.
It is not AI that decides to do up an article for a lazy researcher – that is the choice of the lazy researcher. AI does not decide to stalk workers – that is the decision of controlling managers. AI does not identify the best kinds of workers. Indeed the opposite, has already been found out to be true. Consider the story of pattern matching and algorithms for Amazon hiring when it became evident those algorithms were highly gender-biased, reflecting the (gender-biased) information had been fed into the software. But algorithms for identifying a particular kind of work and workers are just that – algorithms. They are only actualised, they can only change an employee’s world, when someone with enough power and resources decides to use the AI tools. Until then, they are just software/technology/tools.
So let’s limit the spread of myths giving technology human capacities. Let’s unpick and highlight the reification (AKA anthropomorphisation) of AI. Let’s put paid to unfounded expectations of AI, and simplistic (and sometimes incorrect) assumptions about it. AI won’t take jobs. Employers and businesses will decide how the technology is used to subsitute or deskill labour, just as they have shaped and controlled and overseen labour processes and workplace relations for centuries. The imperatives for business have been profits and share prices. These are the drivers that will determine the use of AI, the jobs lost, the surveillance of workers, the possible ’transformations’….
Let’s try to understand AI and the immense power the owners of the technologies have, but let’s not ascribe agency and power to AI. Rather, we must remember that AI needs business and employers to use these latest technologies as tools for the accumulation of profits and power. Employers have long known that new technology is merely a tool – often a very useful one, but still a tool. As scholars seeking to understand labour and work, the economy, sustainability, equality, human rights and social justice, we should all make sure we question the rhetoric critically and avoid the ascription of human agency to AI. We should all campaign. ‘Down with the reification of AI’!
And in this issue?
It has had a gestation that would embarrass an elephant, but it is with great pride that the ELRR editorial team and the wonderful Themed Collection Guest Editors offer you 36(3) of this journal. The first part of this issue is the concluding tranche of the outstanding Themed Collection, Gender and Work – Emerging Issues. The first part of this magnificent Themed Collection was in issue 36(2), although the the first article in this Themed Collection was Promise and peril: Gender, technology, and the future of work in the legal profession by Lee, Foley, Tapsell and Cooper published in 35(4). The Guest Editors Dr Yuvisthi Naidoo and Honorary Associate Professor Anne Junor (together with Dr Tanya Carney who died in September 2024), have done an outstanding job with the dozens of articles submitted. Naidoo and Junor not only ensured peer review was rigorous (thank you to the many excellent expert reviewers but also encouraged, goaded, and aided authors to improve what was already very good scholarship. Finally these guest editors marshalled the final 18 or so articles into appropriate sub-themes. Their guest editorial article introducing Part B of the Themed Collection is a great scholarly overview of the emerging literature and directions of scholarship in this area.
The remaining five general articles in this issue are as broad as ever – testimony yet again to the wonderful multidisciplinarity of this journal.
The first article in this section is Health and Inequality in Australia. As Morris shows, by comparison with most countries, Australia has an admirable and accessible health system, but it is too easy to be satisfied with being better than others rather than excellent. In this wonderful, comprehensive and analytical overview analysis, Morris explores the multiplicity of ways in which the practices, procedures, behaviours and attitudes constrain or prevent access to good and fair health care for many in the community ‘so that class, gender, racial and spatial inequalities are mirrored in the health system’ and become evident in ‘inequalities in life expectancy, health access, utilisation and outcomes’ for disadvantaged groups.
In a significant change of pace, focus and approach, Nyland and Bruce explore the history of the changing nature of trade unionism and the positive links between unionisation and the Taylor Society, an important group of scholars/managers influential in the USA, especially from the 1920s, and which always promoted a participatory agenda rather than an adversarial bargaining agenda for unions. The authors conclude Taylorism, Trade Union Density, and Macroeconomic Stability by showing that ‘in an era characterised by revitalised support for knowledge-intensive reindustrialisation, revisiting the scientific managers’ agenda might assist trade union renewal’.
In another change of focus, pace and place, the article Too Exhausted and Financially Strained to Have Babies? Examining 996 Overwork and Household Digital Finance Among Labourers in China is directed toward the 996 working system – 9 am to 9 pm, 6 days a week. While the 996 system violates China’s labour legislation, it is still widespread. Dong, Guan & Carabetta analysed the panel data from 7,582 family observations in China between 2018 and 2020 and have provided critical theoretical insights into the dynamics of this form of modern slavery for overworked labourers and their families in China. They conclude with rich insights for policymakers and pro-labour organisations.
The focus of Semtner and Dzator is on the nature and impact of soft skills on access to and effective practices in employment. In Soft skills, unemployment risk, and Australian economic downturns, the authors explore how soft skills affected the risk of unemployment from the Global Financial Crisis (GFC) using the Household, Income and Labour Dynamics in Australia (HILDA) They found that the soft skill measures of social capital and low task repetitiveness were associated with lower unemployment, overskilling, and underutilisation risk, albeit at different times after the GFC.
The final research article in this issue comes from Zimbabwe in Africa, where Ndhlovu, Mukuze & Ndhlovu interrogate the Stakeholder perceptions on the retrenchment laws in Zimbabwe. They found that employers tended to avoid using special measures specified in the legislation, while other stakeholders had difficulties with the costly and complex legal processes. Stakeholder perceptions on the retrenchment laws in Zimbabwe also highlight the tension between ‘organisational survival and employee rights, framed through proximity justice and organisational justice theories’.
Overall – as ever – this is an exciting and varied issue which highlights the breadth of scholarship available to readers of The Economic and Labour Relations Review. We hope you enjoy reading the articles within the Themed Collection and in the general Collection – and we hope you become just a bit provoked or goaded by some of the articles!!
Open access
