Hi lovely readers,
If you have spent any time online lately, you have probably seen the endless debates about which programming language you should learn next. Should it be Rust because everyone keeps saying it is the future? Should you stick with C# and .NET? Is JavaScript still worth it, or is it finally time to pick up Python because of all the AI tooling?
These are fair questions, and I love a good language debate as much as the next developer. But after spending the last couple of years watching how AI tools changed the way I actually work day to day, I have come to a slightly different conclusion. The most important language for a developer in the age of AI is not C#, JavaScript, or Rust.
It is English.
You are not writing code anymore, you are describing it
For most of my career, the gap between an idea in my head and working code was filled by me typing that code out, line by line. I knew the syntax, I knew the patterns, and the language I used to get from idea to result was C#.
That gap looks very different now. A large part of my day involves explaining what I want to an AI tool, reviewing what it gives back, and then explaining what was wrong or what I want changed. The bottleneck is no longer "do I know the syntax for this LINQ query." The bottleneck is "can I describe this problem clearly enough that something else can solve it."
That is not a coding skill. That is a communication skill.
If you ask a vague question, you get a vague answer. If you describe the wrong problem, you get a solution to a problem you do not have. The clearer you can express what you actually need, the better the result. And the language most of these tools were trained on, the language most documentation is written, the language most of the prompts and examples online use, is English.
I learned this the hard way in my second language
If you have read some of my other posts, you already know that English is not my first language. I am Dutch, and during university my teachers regularly pointed out how I would lose my train of thought halfway through a sentence. Writing clearly in a second language took me years of effort, which I learned mostly through blogging. That effort turned out to be one of the best investments I ever made, and I did not even realize it at the time.
When you work with AI tools, the quality of your output is tied directly to how clearly you can phrase things. A messy, half-finished thought gives you a messy, half-finished answer. A precise description, with the right context and the right constraints, gives you something you can actually use. The skill of taking a fuzzy idea and turning it into clear sentences is exactly the same skill I had been practicing for years without knowing how useful it would become.
So if English is not your first language either, I want to be honest with you. Yes, this puts a little extra weight on something that is already harder for us. But it is also a skill you can build, the same way you built your ability to read a stack trace or debug a null reference exception. It just takes practice.
Communication was always a huge part
Here is something that I think a lot of us already knew before AI made it impossible to ignore. The hardest parts of software development were rarely the code itself.
The hard parts were understanding what a stakeholder actually wanted instead of what they said they wanted. Writing a pull request description that your reviewer could understand without three follow-up messages. Explaining a technical decision to someone who does not write code. Leaving a comment that your future self would thank you for. Writing documentation that someone could actually follow.
All of that is language. All of that is the ability to take something complicated and make it understandable for another human being. AI did not create this skill, it just turned up the volume on how much it matters. The developers who can clearly explain a problem, give good context, and describe what "done" looks like are the ones getting the most out of these tools right now.
Reading is half of the skill
It is easy to focus only on the part where you describe what you want, but there is another half that matters just as much. You also have to read carefully.
AI tools are confident even when they are wrong. They will happily invent a method that does not exist, suggest a package that was deprecated years ago, or solve your problem in a way that quietly breaks something else. If you cannot read the answer critically and notice when something is off, you are going to have a bad time.
That critical reading is, again, a language skill. It is the ability to follow an explanation, spot the part that does not make sense, and ask a sharp follow-up question. The same way a good code reviewer reads a diff and immediately senses something is wrong, you now need to read generated answers the same way. The better you are at reading, the less likely you are to copy something into production that you did not actually understand.
Programming languages do not stop being important
I want to be clear, because I can already see the comments online. I am not saying you should stop learning programming languages, or that syntax no longer matters, or that you can ship software without understanding what the code does. You absolutely still need to understand your stack. You need to know enough to tell when an answer is nonsense, to debug the thing when it breaks at 4 PM on a Friday, and to make real decisions about architecture and trade-offs.
Programming languages are still the thing the machine runs. But English has quietly become the thing you use to get there. One does not replace the other. They sit next to each other, and right now, a lot of developers are not putting enough effort into the second one.
How to actually get better at it
The good news is that this is a skill you can practice, and you probably already have the perfect tools for it.
Write more. Write blog posts, write detailed pull request descriptions, write proper commit messages, write documentation for the thing you just built. Every time you force yourself to explain something clearly in writing, you are training the exact muscle that makes you better at working with AI.
Pay attention to your prompts. When you get a bad answer, do not just try again with the same vague request. Ask yourself what context was missing, what you assumed the tool already knew, and how you could have phrased it more precisely.
And if English is not your first language, do not let that stop you. Read in English, write in English, and let yourself be a little uncomfortable while you improve. I promise it gets easier, and it pays off in more ways than you expect.
That's a wrap!
Thanks for reading! If you have thoughts, disagree with me, or want to share your own experience, feel free to leave a comment or reach out on my socials. I am always happy to talk about this stuff.
Top comments (0)