AI will not replace you… so far…
We have to admit that all reflections on the topic “AI will soon replace developers” are completely wrong. AI already stepped into our industry with a heavy foot and made a huge impact on the hiring process. It is not only about “replacing”, but also how much fresh blood and energy will be lost by replacing promising candidates with AI assistants.
When a developer creates a unit test file and the copilot produces 80% of all valuable unit tests for the related class — it is already an example of a huge impact to all eco-system around because eliminates the need to delegate some portion of the job to interns and junior developers, so the chain of knowledge sharing is already broken.
You don’t need to spend your time onboarding new engineers, you don’t need to build proper habits and culture — there is already efficiently trained invisible co-worked near you that follows your design styles and produces code increment that in most cases (after some period of adoption) compatible with your expectations.
But not only interns and juniors positions are affected. When a front-end developer safely creates a Jenkins file to define a project CI — there is somewhere DevOps crying. From this moment all this internal, private, secretive knowledge that was owned by a specific department is no longer a private asset. Everyone with basic skills in a specific technology area can obtain superpowers just by delegating AI to make the first step.
When asking ChatGPT about nuances in implementation inside specific technology (e.g. injecting Vault secrets inside HTML static page hosted inside Nginx orchestrated by Kubernetes) and obtaining answers in seconds eliminates the need for meetings that can take hours based on the number of departments involved. Even having an answer not 100% accurate gives you a good understanding of how to move forward, where particularly dive deeper.
As a result, using AI as a permanent assistant makes a single developer full-sufficient and comparable in performance with a full team of engineers. And exact this superpower dramatically affects the hiring process and reduces pressure on HR departments with requests to find specific engineers with narrow but strong skill sets. We don’t need “a team” to build complex projects anymore — all this concept of “2 pizza” team will not be relevant very soon.
And as a consequence, the arguing on the topic of what is better “generalist” or “specialist” is no longer relevant as well — with the help of AI assistants everybody can be and generalist and specialist at the same time. But, to be honest, being a “specialist” is already to be in a dangerous space — you have to compete with AI so have small chances to win this competition.
AI revolution is changing IT landscape dramatically, forcing engineers to become more versatile and self-sufficient in their fields, forcing to build good abstract problem-solving skills rather than become good technology expert, eliminating need to be part of strong team. Increased performance threshold that already includes AI assistant steroids is also new reality for tech companies. Are we ready to accept new rules?
