My neighbour, who has no software development experience, just built a financial services platform using Claude.
Think about that for a moment.
Someone who would previously have needed a developer to build their idea can now sit down with an AI assistant and build it themselves.
Now imagine that same person looking at an AI engineering job posting. They have an application to show. They know how to use Claude. And they're willing to do the job for 50% less than your salary.
How long before they apply?
More people will be able to do the work
Software development used to require enough technical knowledge to turn an idea into code. That requirement kept a lot of people out, even when they understood the problem they wanted to solve.
AI is changing who can participate. Someone with experience in finance, operations, or sales can describe a problem in language they already understand and ask an assistant to build a solution.
They can try it, ask for changes, and keep going. The cost of getting started is much lower than it used to be.
As the tools improve, I expect more people without technical backgrounds to take on work that companies currently hire engineers to do.
What happens when the interview allows AI too?
If AI becomes fully integrated into software development and technical interviews, the same tools helping engineers at work will also help new candidates compete for those jobs.
A person who couldn't have completed a coding assessment alone may be able to complete an AI-assisted one. Give them a take-home assignment, access to Claude, and time to work through it, and they have a chance to produce something worth evaluating.
Their application no longer consists of enthusiasm and a promise to learn. They can show you what they built.
That doesn't mean every AI-built application is good. Candidates still have to demonstrate that they understand and can maintain their work. But experienced engineers should take seriously how much easier it is becoming for someone else to enter the competition.
They'll call themselves AI Engineers
Sooner or later, I expect companies to receive more applications from non-technical people calling themselves AI Engineers.
Some will come from the very industries those companies serve. They'll understand the business problem, have an AI-built application in their portfolio, and be prepared to accept much less than an experienced engineer earns.
Picture the conversation: "I've already built something similar. I can use the same tools your team uses. And I'll do the job for half the salary."
A company trying to control costs will have a reason to look at that candidate.
You may see the gaps in their technical experience immediately. A hiring manager may also see a lower-cost candidate who can deliver some of the work the team needs.
The salary pressure doesn't have to wait until that person can do everything you can do. It can start when an employer believes they can do enough.
My neighbour built a financial services platform without a software development background. That's what made me think about where this is heading.
If more people can build with AI, and more interviews allow AI, the pool of people competing for engineering jobs will grow. Some of them will be willing to do the work for 50% less than you.
Comments