Generative AI v/s Traditional Programming: An Unbiased Look In 2023

Updated on November 7, 2023

Traditional programming, at least the way we know it, is dead.

Nope, we were not the first one’s to say it.

We read it on Dev.to, you know where all THE programmers hang out.

The internet is abound with videos and articles on how coding as a job is dead. After some digging around, we found out that there is some truth behind this statement.

But it’s not all black and white. There are gray areas here, and it is our job to look at this story from both sides.

If you don’t want to read the entire story, here is a quick TL:DR

• Software engineering and coding has been around since the 40s.
• There have been pivotal moments in the history of programming, and the introduction of ChatGPT was one such point. But will ChatGPT and other LLMs make coders obsolete?
• LLMs like Google Bard and ChatGPT may not replace programmers, but they will change the programming profession by making programmers more efficient.
• Will we see programmers being completely replaced one day by AI?

So, let’s dive right into Generative AI v/s Traditional Programming.

How software development evolved

Before we see where software development is going, let us take a quick look at how we got here. Don’t worry, we won’t bury you in a lot of history here.

Let this infographic do the talking:

So writing software has always been a “human” thing, where you told the machine what to do and the machine did it, like a loyal dog.

As you can see from the infographic above, the field evolved, and is still evolving, fast. Over the past three decades, especially since the introduction of personal computers, open source programming and smartphones, traditional programming has undergone sea changes.

And the launch of ChatGPT at the tail end of 2022 was a pivotal point in this evolution.

Let us now see how LLMs like ChatGPT are changing the world of programming.

ChatGPT v/s Programmers: Coders made AI, and AI will destroy Coders?

Before we talk about ChatGPT, the evolution of programming deserves one honorable mention.

GitHub’s Copilot.

CoPilot came into existence a good one year before ChatGPT, in October 2021, to be precise.

GitHub’s CoPilot made many people feel like Gilfoyle. Coders were suddenly getting tasks done with 10x the speed.

A study by GitHub of more than 2000 developers, in fact, found that Copilot cuts the coding time by nearly half for certain development tasks. And this is not just auto-completion of code that we are talking about.

GitHub Copilot made a huge impact on the way programmers wrote code, but there are still ambiguities in natural language. Microsoft stepped in to solve these ambiguities with the introduction of TiCoder, which claims an improvement from 48 percent to 85 percent.

But it is ChatGPT that is changing the way software development is being done.

ChatGPT 4 was multimodal, in the sense it took in more than text prompts. ChatGPT 4 can process images, and, in a livestream organized by OpenAI, Greg Brockman, one of OpenAI’s leading members, showed how powerful it actually was.

We have written about this earlier, but the gist is..

Soon, coders were using ChatGPT to get their jobs done faster, and a plethora of AI tools flooded the market, all promising to save coders precious time (and most of them did).

But the original LLM aka ChatGPT, still  rules the roost. For instance, Ammaar Reshi a UX designer from England, but he built a simple Snake game using ChatGPT and Replin, all within a matter of 20 minutes.

ChatGPT can’t make programmers obsolete

A lot of programmers agree that while ChatGPT can produce code snippets of code, it can never fully replace a full fledged software programmer. This is because programming is a multidisciplinary function that involves a combination of logical thinking, imagination and mathematics, among others.

An LLM is trained on datasets, but it cannot match the neural network of the human brain.

This is because while ChatGPT provides you with the exact code needed to get the job done, it does leave blanks for you to fill. So, say you want to write the code for an app that does accounting, you need to know how accounting works, and, at the same time, be familiar with the programming language that you can use to code the app.

But LLMs can still make programmers more efficient

Yup, LLMs are here to stay, and, like all professions, Generative AI tools like ChatGPT will change the way programming is done.

Here are a few ways the programming profession can benefit from tools like ChatGPT:

1. Cleaner code:

ChatGPT can generate code that uses the CPU more efficiently, and also helps in tracking security vulnerabilities. This will help developers write more efficient code.

2. Easy debugging:

 One of the most tedious tasks in programming is debugging, and in many cases, finding the bug is more difficult than writing actual code. ChatGPT not only helps developers find solutions to error messages faster, it can also fix the bug and explain the fix in some cases. Scouring the web and developer blogs looking for a solution will now become a thing of the past.

3. Understanding code:

ChatGPT can explain in simple terms, complex pieces of code. This will save countless hours for developers, who don’t have to spend time looking at the code’s underlying logic.

4. Automating testing

ChatGPT can be helpful for developers to write automated test codes. Developers thus have more time to write code than testing the code for bugs and errors.

As you can see, LLMs like ChatGPT can make developers more efficient, making them work smarter and saving precious time.

But will AI one day completely upend traditional programming?

We look at it in the final section of our post.

Will AI one day replace programmers entirely?

If you have read this far, then you know that the answer to this question is NO.

“I don’t believe AI is anywhere near replacing human developers. It will remove the mundane, boilerplate stuff that people have to do, and they can focus on higher-value things.”

—Vasi Philomin, Amazon

That being said, we believe that these LLMs will get more and more efficient as time goes by, making programming accessible to a much larger audience.

Low code no code platforms  can now be easily built, and novice programmers who are just familiar with the basics of a programming language will soon have to level up their game.

Human beings will always be in need to ensure that the machine understands the intent with which the code was written. And, if the machines one day do get “Smarter” than programmers one day, with the rise of Artificial General Intelligence, then worrying about your programming being done by a machine may be on the bottom of the list of things you need to worry about.

---

At Kommunicate, we envision a world-beating customer support solution to empower the new era of customer support. We would love to have you on board to have a first-hand experience of Kommunicate. You can signup here and start delighting your customers right away.