After all, the people getting really rich were the software folk: Atlassian founders Scott Farquhar and Mike Cannon Brookes; Canva’s Melanie Perkins and Cliff Obrecht

Stories abounded of schoolchildren dropping out and selling apps and computer games for millions of dollars

Sydney Grammar claimed Microsoft and Apple would be the real winners from the then Rudd–Gillard government’s $2

4bn Digital Education Revolution program to fund laptops for students, and the various federal and state policies encouraging them to learn how to code

Today, as tens of thousands of coders lose their jobs, that prediction looks right

On Thursday Atlassian said it would fire 1600 staff, or roughly 10 per cent of its global workforce, citing a fundamental “rebalance” toward artificial intelligence

Experts were surprised that number wasn’t higher

The decision marks a grim milestone for a generation of young Australians who were told that learning to code was the ultimate insurance policy against the vagaries of the modern economy

For a decade, politicians and educational institutions have relentlessly championed “computational thinking” as the new literacy, effectively promising students that a career in software engineering was a golden ticket

Queensland Premier David Crisafulli is still backing the bet, this week announcing his government’s 2026 Premier’s Coding Challenge, “to inspire Queensland students to use and build their coding skills”

The irony is that the original academic definition of “computational thinking” was about logic and problem solving

Unfortunately it was repackaged to Australian parents as “learning to code” for job security

The opposite has proved true

Earlier this month Afterpay owner Block announced that half its staff, equating to 4000 jobs, would go as a result of AI and, just a week before that, WiseTech Global cut 2000 jobs, citing the same reason

Now that AI can handle the coding part, the “literacy” students were taught (syntax and languages) is becoming obsolete, while the “thinking” part (logic and architecture) is what they actually needed

Australian National University honorary associate professor Matthew Darling says coding would be a brave career choice at the moment

“From a career perspective, it would be courageous to enrol in programming courses right now,” says Darling, who created an award-winning rapid coding system before moving into academia

For the thousands of Australian students currently enrolled in computer studies or coding bootcamps, the news probably feels like a governmental betrayal

The narrative that every Australian child should be a programmer began in earnest around 2015

At that time, the political class was gripped by a “STEM” fever, arguing that traditional subjects like history and geography were secondary to the digital languages of the future

In his 2015 budget reply speech, then-opposition leader Bill Shorten declared that “coding is the literacy of the 21st century”

“Digital technologies, computer science and coding … should be taught in every primary and every secondary school in Australia,” Mr Shorten said

“Under Labor, every young Australian will have the chance to read, write and work with the global language of the digital age

” The Coalition was equally bullish

The then education minister Christopher Pyne, while introducing a revised national curriculum, said Australia would “fail as a country” if it did not pivot immediately toward code

When Malcolm Turnbull took the prime ministership, he launched the $1

1bn National Innovation and Science Agenda (NISA)

The “innovation prime minister” frequently exhorted Australians to be “as imaginative as we can” and poured $51m specifically into programs for primary students to learn coding

The white papers of the time, such as the Shape of the Australian Curriculum: Technologies, codified this shift

They argued that students must be “creators of digital solutions” rather than mere “consumers”, embedding programming into the syllabus from as early as foundation level

For today’s cohort of junior developers – many of whom began their schooling just as these policies took effect – the Atlassian cuts suggest they may have been sold a pup

The promise was simple: learn to write code and you will always have a high-paying job

But the rise of generative AI, and particularly Anthropic’s Claude, has changed the maths

Large language models can now generate, debug and optimise code in seconds – tasks that previously required armies of entry-level engineers

Universities and vocational institutions, which spent the past decade expanding their IT departments to meet government-stoked coding demand, may now face an identity crisis

Darling points out though that computer science, which looks at the fundamental building blocks of computing such as semiconductors, quantum computing, advanced operating systems and algorithms, is still a “vitally needed area of skill”, as is software engineering

He also points out that for all the talk of creating code being as easy as punching in a prompt in Claude, the user would need to know what they are asking for

“You’ve got to be able to say what you want it to produce, from what I would term the prism of technical feasibility

There’s no point in asking AI to build something that is impossible to build,” he says

“For instance, how do you contain it? How do you control it? How do you get the most out of it and maintain the mental acuity that you need to be able to drive it effectively?” Nevertheless, the actual job market is signalling that “computational thinking” may be a commodity right now, not a career

For 10 years, politicians and many schools promised that the future belonged to the programmers

Today, as the founders of our most successful tech export “rebalance” their way into the AI era, that promise looks increasingly hollow