Is Generative AI already eating software jobs?
Many a pundit have pointed to the declining headcount numbers of prominent Indian IT players in recent quarters and attributed it, at least partly, to the rise of GenAI and its ability to generate codes. Some have highlighted the recent job cuts at Big Tech companies and concluded that GenAI is at work. Who needs software developers now?
A billionaire CEO and founder said recently that prompts will replace programming, and possibly the need for programmers.
GenAI is indeed quite a force, and of course, the most buzzing topic of the past 18 months.
But does it spell doom for software engineers?
***
Let's look at calculators - the most basic computing tool.
They have been around for decades. Many jobs that in the olden days required back-breaking manual computations (e.g., interest and tax calculation, invoice reconciliation) are no longer done manually, thanks to calculators (and of course computers in more recent times).
Yet, we have not stopped teaching basic math to our kids. Why?
Because we still need to conceptually understand what calculation needs to be done in a specific situation before we assign the computation part to the calculator/computer.
If a simple calculator seems too simplistic, try visualizing a scientific calculator and imagine operating it without a good understanding of the concepts underlying the calculations that it is capable of doing.
This I think is a good analogy to hold in mind while evaluating code writing capabilities of GenAI and it's implications for us.
***
To generate useful codes using GenAI, you still need to understand well the basics of programming - language fundamentals, data structures, algorithms, and so on. Perhaps more so than in the past, because If you don't, you can end up generating a lot of crappy code at scale.
Also, being a citizen coder and developing simple applications using GenAI is one thing (and quite a fascinating thing, I must admit), but developing enterprise grade software is a completely different ball game.
Code writing is just one part in the software development process. Understanding user needs, defining problem statements, designing systems, deciding on the tech stack, testing, deployment, maintenance are all equally important in the process - and can not be done by GenAI alone.
***
Jury is still out on whether GenAI will replace software developers.
I think it won't.
It will only weed out the weaker ones, but will likely multiply the demand for good ones.
Two things every software developer must focus on to thrive in the era of GenAI:
1. Domain: a deep understanding of the domain/industry. This will help you better understand business needs, define the problem statements and outline relevant solutions.
2. Back to basics: a solid grasp on the basics of programming - language fundamentals, data structures, algorithms, and so on.
#genai #softwaredevelopment #jobs #itcareers