In today’s tech landscape, the ability to understand and work with code that isn’t your own is more crucial than ever. Why, you ask? Well, let me break it down: Large Language Models (LLMs), the super-smart AI systems, now make generating code a breeze. As a Technology Consultant working with various companies, I’ve noticed many folks turning to tools like GitHub Copilot or ChatGPT to simplify coding. It’s convenient and cost-effective, but there’s a twist.

[

View this post on Instagram

](https://www.instagram.com/p/Cyc451jOHF3/?utm_source=ig_embed&utm_campaign=loading)

A post shared by Ashok Tankala | Leadership and Executive Coach | Ex-CTO (@ashoktankala)

The Mystery of Unfamiliar Code

Dealing with code you didn’t write can be like solving a puzzle. If you don’t understand it, you’re in a tricky spot. Bugs might pop up, or the code could act oddly. Plus, there’s a cybersecurity angle. Hackers could train LLMs to create code with hidden security issues. If that happens, it’s a problem not just for you but also for your company.

The Rise of Test-Driven Development (TDD)

In this new era, Test-Driven Development (TDD) becomes a superhero! It’s a way to ensure that the code works correctly. While we can ask LLMs for tests, relying solely on them is risky business. If the LLM makes a coding mistake, the tests it generates will be flawed, leading to low-quality work.

The Code-TDD Partnership

When LLMs give us code, we can use TDD to double-check its quality. Instead of crossing your fingers and hoping for the best, we kick things off by writing or tweaking tests. This guarantees top-notch code without sneaky errors.

Your Thoughts, Please! What are your thoughts on the age of LLMs?

The post Exploring the Role of Code Understanding in the LLM Era – A Tech Consultant’s Perspective first appeared on A Software Engineer's Journal.