Last week, I had a brief conversation with a former student who had attended many of my training sessions. While discussing future career growth opportunities, a question that has been making the rounds across tech circles came up: “With AI improving every month through accelerating innovation, how long before it replaces our jobs? What will happen to software engineers when AI systems can write code, test, refactor, and build software at a scale and speed beyond human capacity — at a fraction of the cost?”
These questions are keeping the current generation of software engineers restless, casting a shadow of despair over their future career prospects. The wave of layoffs sweeping the tech industry today may well be an early indicator of what lies ahead.
And this uncertainty won’t stop at software. It is poised to ripple outward — into financial services, journalism, entertainment, healthcare, the legal profession, and beyond. Which begs an uncomfortable question: when large corporations inevitably embrace AI and automation to cut costs and boost profitability — replacing human labour with machines that think and act — how will society cope with this disruption? Could this spiral into a doomsday scenario for humanity?
Lessons from History
To understand what may come, it helps to look at what has already passed.
Just over a century ago, urban transport across the world ran on horsepower — literally. The bustling city streets of Europe in the late 1800s and early 1900s were crowded with riders on horseback and families travelling in horse-drawn carriages. Then the automotive industry took off in the 1920s, and within a single decade, horses and carriages had largely vanished from city roads, replaced by cars and motorcycles.
What became of the horse riders, farriers, carriage makers, hay suppliers, blacksmiths, and stable hands? They adapted. Horse riders became taxi drivers. Farriers became tyre specialists — patching punctures and maintaining pressures. Carriage makers reinvented themselves as auto body craftsmen, skilled in repainting, dent removal, and interior work. Hay suppliers became oil suppliers and gas stations. Blacksmiths became car mechanics. Horse stables became garages and car showrooms.
The jobs did not simply vanish — they transformed.
The Keypunch Generation
Consider another parallel, closer to the world of computing. In the 1940s and 1950s, at the dawn of the computing age, many people — predominantly women — were employed as keypunch operators. Their job was to use keypunch machines to encode data onto punched cards, which were then fed in batches into mainframe computers through card readers for processing.
These keypunch machines resembled mechanical typewriters, except that instead of printing on paper, they punched small rectangular holes at specific positions on an 80-column punch card. It was precise, skilled, and demanding work.
As computers evolved from early-generation mainframes to more sophisticated minicomputers with interactive time-sharing systems, punched cards and keypunch machines were rendered obsolete, replaced by teletype terminals (TTYs). Yet the people behind those machines did not simply disappear — they evolved. Punched card operators became systems administrators. Keypunch operators became programmers, testers, and QA engineers.
The Coming Wave
The coming wave of AI and automation will likely follow a similar trajectory.
AI will certainly eliminate some categories of work — but the larger shift is not simply about replacement. It is about the redistribution of cognitive labour. As automation becomes cheaper, faster, and more capable, the economic value of purely repetitive execution begins to decline. Human effort gradually migrates toward areas involving supervision, intent, interpretation, coordination, trust, creativity, and systems-level understanding.
New forms of work will emerge around these transitions:
- robotic systems maintenance and oversight
- AI auditing and alignment engineering
- human-AI workflow orchestration
- synthetic media verification
- domain-specific AI supervision
- intent engineering
- and entirely new professions we cannot yet clearly predict
This pattern has repeated throughout technological history. The natural rhythm of human life — retirement — will also play its part in smoothing this transition, as older generations phase out and younger ones pivot toward the opportunities that emerge. A software engineer with three or more decades of experience will have already witnessed several such shifts. In the 1990s, there was a wave of Visual Basic and Delphi programmers, database engineers working with FoxPro, Clipper, and Microsoft Access, and COBOL developers racing to fix Y2K issues. Most of those roles have since vanished or transformed — into web developers, J2EE architects, and cloud specialists of the 2000s and beyond.
Every decade in the software industry has brought a significant trend shift: those who adapt survive; those who don’t, eventually retire.
The Deeper Shift Beneath Automation
But there is an even deeper pattern beneath these transitions.
Every major technological leap changes what society itself can coordinate. Railways expanded the scale of logistics. Telecommunications compressed distance. Computers accelerated information processing. The internet transformed global coordination. AI is now beginning to compress cognitive execution itself.
That distinction matters.
When a technology consistently reduces cost, latency, effort, or coordination overhead, society gradually reorganises around it. What begins as a competitive advantage slowly becomes infrastructure — and eventually, part of the environment itself. This is why AI adoption may not primarily occur because organisations want to adopt it. It may occur because competitive pressures make non-adoption increasingly difficult.
Faster execution becomes expected. Lower operational cost becomes expected. Continuous optimisation becomes expected. The surrounding environment shifts, and institutions adapt to survive within it.
Why Learning Programming Still Matters
This shifting landscape is precisely why younger generations increasingly question the relevance of learning programming, algorithms, and data structures in an era where AI can generate functional code from a natural language prompt. If the machine can write the code, why learn to write it yourself?
But the value of learning programming was never merely about typing syntax into a machine.
Consider arithmetic. For decades, humanity has relied on calculators for everyday computation. Yet children still learn arithmetic manually in school — they memorise multiplication tables even though a calculator can produce the answer instantly. Why? Because understanding matters. We learn arithmetic not because humans can outperform calculators at computation, but because mathematical understanding helps us reason about problems, validate outputs, recognise errors, and develop intuition about the systems we interact with.
Programming increasingly occupies a similar role. Learning software engineering develops structured thinking, decomposition of complexity, logical reasoning, abstraction handling, and the ability to communicate intent clearly. These capabilities remain valuable even as machines increasingly handle the implementation details.
Intent Versus Execution
AI systems are extraordinarily powerful at execution. They can synthesise information rapidly, generate code, automate workflows, analyse patterns, and scale tasks far beyond individual human capacity.
But human intelligence operates differently.
Human cognition is deeply tied to intent, curiosity, judgment, meaning, ethics, lived experience, and contextual understanding. AI may optimise for a goal — but humans still define why that goal matters in the first place. In this sense, AI resembles an externalisation of intelligence into tools and infrastructure: an amplification layer built upon human civilisation.
A well-crafted sword can be extraordinarily sharp. But the sword itself possesses neither intent nor wisdom. Its consequence depends entirely on the one who wields it.
A Signal to the Next Generation
This impending shift is already sending a subtle but clear signal to those entering the workforce: choose your career path wisely. Invest in skills that will remain valuable in an increasingly automated world — skills rooted in judgment, creativity, communication, and above all, intent.
The next generation may need to think less in terms of static professions and more in terms of evolving capabilities. Because throughout history, technology has rarely ended human relevance — but it has repeatedly transformed what society considers a valuable human contribution.
The machines are getting sharper. The question is whether we are sharpening ourselves to wield them well.