The Developer Skills gap in the age of AI

Hello everyone, I hope you are doing well. In this blog, we will deep dive to address a very important question which is how will the skill gap be affected in the age of AI. You see with the rise of tools like GPT, Gemini, or Devin, we have seen videos of the sort that "AI will be your tutor" or "You just need to get started and AI will handle it".

Yes while that's true that you can ask AI to build components, and get in-depth explanation about topics you don't know about, it is also true that AI is hence making new developers dependent on it and not applying their skills.

To prove this, I recently met a junior developer who was just getting started with Database and DB Tools. Hence I advised him to learn first about the concepts of Databases like Joins, Relationships, ER diagrams, etc., and then proceed to learn about tools like MySQL, and MongoDB and how to connect them with your application.

My man directly jumped to learning MongoDB and Prisma and how to set them up. His idea was that he would learn the concepts on the go if needed or might not learn at all. Later, what happened was that he could not start his project because he was stuck in declaring many-to-many and one-to-one relationships in Prisma the way he thought. He asked GPT with prompts but was not able to go through. Ultimately he had to start learning again.

Now yes one can argue the fact that he could have made better prompts to get what he wanted and yes that's true but that's not the point! The point I am stressing is that now in the age of AI, developers will need to focus on the basics very very carefully.

In this example only, say the junior developer had learned and cleared all the basic concepts of the database first, then even if given the wrong code by AI, he could have just modified it quickly and continued with the project but now his project will be delayed. In the future, products and services will be shipped faster than ever due to various obstacles being handled by AI tools and hence there must be minimal delay in building the product.

If you know the concepts (theory in boring terms) well, you will exactly know what you want from AI and can also modify quickly if needed. You can see the proof by watching the live streams of several cool developers.

Hence in the age of AI, the real skill gap will be the absence of knowledge of theory and concepts that drive the domain. Everything else can be obtained directly or indirectly.