AI Killed IT Elitism

For decades, IT was the go-to industry if you wanted a high-paying job without going to medical school or chasing a corner office. Designers, testers, project managers, all well paid. But the real elite, the rock stars of the office floor, were the developers.
Sure, top executives and brain surgeons earned more. But among normal salaried jobs, software developers sat firmly at the top of the pile.
I say "sat", because that has changed. Completely, and I'd argue irreversibly.
With models like Claude Opus 4.7 and GPT-5.5, and tools like Claude Code and OpenAI Codex, solving day-to-day problems and building small tools to automate work has become a basic professional skill. Not a specialist one. Basic.
Today, when hiring for any office role, you assume the candidate can use a computer. Within a year or two, the same will apply to AI. Every professional will be expected to brief a model, get useful output, and stitch together a small tool for repetitive work. Anyone who can't will look the way someone who refuses to touch a computer looks now.
And that means the skill once reserved for the "gifted few", writing code, has become ordinary. AI writes code faster than any developer, and with the latest model generation, often better. The mystique is gone.
Building software is now like cooking
Want a quick bite? Stick a pizza in the microwave. Want a quick automation? Ask an AI to write you a script.
Want something more substantial? Get the chopping board out and cook properly. Want a robust internal tool? Fire up Claude Code and build an app.
But what if you need a sit-down dinner for a hundred guests at a wedding?
You hire caterers.
And here's the thing about caterers: you're not really paying them to cook. You're paying for logistics, food safety, scheduling, dietary management, hiring waitstaff, equipment, contingency planning. The cooking is a fraction of the job.
It's the same with IT now. When code practically writes itself, the cooking part becomes the cheapest part of the project. What you actually pay a consultancy for is everything around it: architecture, infrastructure, security, integration with the rest of your business, prioritising what to build (and what not to build), and the experience to know which decisions you'll regret in two years.
What this means for the industry
The shift is already happening, and it's brutal.
Developers whose main skill was Googling Stack Overflow until something compiled are going to struggle. So will the body-shop consultancies whose business model was renting out junior coders by the hour without any serious architectural or strategic expertise. There's no margin in selling cheap labour when AI is cheaper, faster and available at 3am on a Sunday.
What about the firms that do have proper expertise? They don't disappear, but their position changes. Software development stops being a glamorous, mysterious profession that justifies six-figure salaries for typing curly braces, and becomes a normal trade. Skilled, valuable, well-paid for genuine expertise, but no longer special simply because you can write a for-loop.
Honestly, it's about time. The elitism in IT was always slightly absurd. The real value was never in the typing. It was in understanding the problem, designing the solution, and shipping something that actually works for the business. AI hasn't taken any of that away. It's just stripped away the bit that was making everyone overpaid.
The elite status doesn't vanish. It moves: from the people who can write code, to the people who know what to build, why, and how to make it last.
