Atlassian layoffs reopen the debate: should AI cut working hours instead of headcount?

After Atlassian cut 10% of its staff, the debate over AI and employment has become far more concrete. Economist John Quiggin writes that productivity gains from tools like Claude should not automatically lead to layoffs and higher corporate profits. The alternative is a return to the idea of a shorter working week, as happened after earlier technological shifts.

AI-processed from Guardian; edited by Hamidun News
Atlassian layoffs reopen the debate: should AI cut working hours instead of headcount?
Source: Guardian. Collage: Hamidun News.
Listen to article
Article Fact Sheet & Key Figures
ParameterDetails
Primary sourceGuardian ↗
Editor-in-Chief
Key metric #1AI already delivers measurable productivity gains in software development.
Key metric #2Without a new balance of power, those gains easily turn into layoffs and higher margins.
Key metric #3Remote work proved that the familiar five-day schedule is not set in stone.

After Atlassian's decision to cut 10% of its employees, the conversation about how AI is changing the labor market is no longer an abstract argument about a distant future. As developers report sharp productivity gains, a question is getting louder: why should only companies capture the benefits of automation, and not the workers themselves?

Why the debate is back

The latest wave of discussion was triggered by layoffs at Atlassian, one of the most prominent players in Australia's software industry. For the author of the column, this is an important signal: the conversation about AI and jobs can no longer be conducted in the conditional mood.

In the software industry, the effect is already visible in practice. Developers report serious productivity gains thanks to tools like Anthropic's Claude, and this changes not only how fast code gets written but also managers' calculations of how many people a team needs.

The problem, in economist John Quiggin's view, is that the debate is usually framed too narrowly. If AI lets people complete the same volume of tasks with less effort, business almost automatically considers two scenarios: increase profit or cut staff. A third option is discussed far less often: reduce working time while keeping people employed. And this option, the author argues, has historically been society's normal response to rising productivity.

"AI discussions seem to assume in advance that if the total number of

working hours falls, someone must necessarily lose their job."

How hours were cut before

Quiggin reminds readers that anxiety over new technology is nothing new. During the Industrial Revolution, mechanization at first did not free people but, on the contrary, made work harder: familiar forms of employment disappeared, and for many the working week reached almost 70 hours. The Luddite protests were a reaction not to abstract progress but to a very real deterioration in living and working conditions.

Later the balance did begin to shift. In the second half of the 19th century and in the 20th century, the benefits of technological progress came to be expressed not only in higher output and business income but also in a gradual reduction of working time. Australia and New Zealand were among the first to push for the eight-hour day, after which the standard week shrank from 48 to 44 and then to 40 hours. Even the weekend we take for granted today was once the result of a long struggle by workers, unions and the state, not a gift from employers.

However, this process stalled. In Australia, a move to a 35-hour week was discussed as far back as the 1980s, but it went no further. The working week settled at 38 hours, and the idea of cutting time further came to be seen not as the next logical step but as something exotic. That is why today's conversations about AI quickly slide into talk of layoffs: the current norm seems natural and unchangeable.

Who gets the benefit

The author ties this debate not only to AI but also to the experience of recent years. The pandemic showed that working arrangements can be changed much faster than previously thought: millions of people moved to remote work almost instantly, and the system did not fall apart. What is more, after lockdowns ended, employers were unable to return the market fully to the old rules. For office workers, this proved that the organization of work is a political and managerial choice, not a law of nature.

There are other signs of a shift. Australia has introduced a right to disconnect from work messages outside standard hours, and some employees have begun quietly, without official announcements, to win back Friday half-days. But these changes are distributed unevenly: those whose work requires physical presence, in retail, logistics and transport, gain far less from the new flexibility.

  • AI already delivers measurable productivity gains in software development.
  • Without a new balance of power, those gains easily turn into layoffs and higher margins.
  • Remote work proved that the familiar five-day schedule is not set in stone.
  • The right to disconnect became an attempt to protect personal time from the "creeping" expansion of the working day.
  • A shorter week will not appear on its own; it takes unions, policy and negotiation.

Against this backdrop, Quiggin criticizes governments that are not ready to discuss a four-day week seriously and often, on the contrary, try to restore tighter control over office attendance. His main thesis is simple: if AI really does make workers more productive, society should argue not only about how many people will be left without jobs, but also about how to divide the freed-up time.

What this means

The Atlassian layoffs have become a convenient focal point for the larger debate over AI and employment. If productivity growth is not converted into a shorter working week or other guarantees for employees, the technological gain will once again settle into company profits, and talk of "efficiency" will mean, for many, only fewer colleagues and more uncertainty.

Story

Key Questions Answered

Why the debate is back?

The latest wave of discussion was triggered by layoffs at Atlassian, one of the most prominent players in Australia's software industry. For the author of the column, this is an important signal: the conversation about AI and jobs can no longer be conducted in the conditional mood. In the software industry, the effect is already visible in practice. Developers report serious productivity gains thanks to tools like Anthropic's Claude, and this changes not only how fast code gets written but also managers' calculations of how many people a team needs. The problem, in economis

How hours were cut before?

Quiggin reminds readers that anxiety over new technology is nothing new. During the Industrial Revolution, mechanization at first did not free people but, on the contrary, made work harder: familiar forms of employment disappeared, and for many the working week reached almost 70 hours. The Luddite protests were a reaction not to abstract progress but to a very real deterioration in living and working conditions. Later the balance did begin to shift. In the second half of the 19th century and in the 20th century, the benefits of technological progress came to be expressed no

Who gets the benefit?

The author ties this debate not only to AI but also to the experience of recent years. The pandemic showed that working arrangements can be changed much faster than previously thought: millions of people moved to remote work almost instantly, and the system did not fall apart. What is more, after lockdowns ended, employers were unable to return the market fully to the old rules. For office workers, this proved that the organization of work is a political and managerial choice, not a law of nature. There are other signs of a shift. Australia has introduced a right to disconn

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Unified AI Workspace · Nomad

Test Frontier LLMs & Autonomous Agents in One Workspace

Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.

Try Nomad AI →Launch Nomad Agent
What do you think?
Loading comments…