Only if you’re looking for an excuse.
No, you should not. Not if it’s your passion, anyway.
Here’s the thing. We’ve been in similar situations before. Don’t believe me? The title of this article is only one word different from another article I wrote: Does All This Talk of Outsourcing Mean I Should Choose a Different Career?
I wrote that article just over twenty years ago.
Everything old is new again.
Coding still counts
You don’t need to give up on a tech career because AI is changing things. Learning to code still helps you think clearly and solve problems. If you know how to code, you’ll also be better at using AI tools.
What coding is and is not
In that older article, I discussed passion, and that hasn’t changed. If tech doesn’t speak to you, then you probably want to choose a different career, no matter what happens with coding or AI. On the other hand, if, like me, a love of technology gets you up in the morning, read on.
Coding still matters. In fact, there’s a chance it matters more than ever.
Coding is not learning a programming language. I mean, a programming language is generally used to teach coding, and you need to learn that language. But in reality, what we call coding is more about thinking. Coding involves the ability to:
- Break down a problem into manageable chunks.
- Understanding when a problem is poorly or incompletely defined.
- Devise multiple solutions to a problem and weigh the pros and cons of those solutions.
- Describe a solution in discrete, clear, unambiguous steps.
- Experiment to discover which solutions work best.
- Analyze your work objectively. Find and correct mistakes, particularly when things don’t behave as planned.
- Know which parts of the solution are more important than others in terms of performance and functionality.
- Think sequentially and logically.
And, yeah, coding is about translating the solutions you have in your head into instructions in a specific computer programming language. Heck, it can even involve selecting which language is most appropriate or knowing when to change languages mid-stream.
But programming languages come and go.1 The skills I’ve enumerated apply to them all.
Help keep it going by becoming a Patron.
Knowing A makes you better at B
This is an opinion I’ve long held that I don’t see articulated often. At the risk of having to use specific programming languages to make my point:
- Understanding hardware and machine code makes you a better assembly language programmer.
- Understanding assembly language makes you a better C or C++2 programmer.
- Understanding C or C++ makes you a better Python programmer.
- And so on.
When you work with a programming language at a particular level of abstraction, and you also understand the level of abstraction just below that, you have a better handle on what’s happening at your chosen level.
Which finally brings us to AI.
Knowing how to code makes you better at AI
Vibe-coding is just a higher layer of abstraction. We’ve just turned the programming “language” into English.
People who understand how to code are better vibe coders. Vibe-coded projects created by people who understand the underlying principles of computer programming are more reliable. They are likely to be produced more quickly and with fewer issues than those with AI-only experience. There’s a reason many coders who were laid off because AI could do their job are being hired back.
All of the skills I listed related to coding make you better at vibe-coding. The better you understand the code produced by AI coding tools — the abstraction layer “one level down” — the more you’re able to define, diagnose, and tweak the results produced by AI.
And you are certainly able to do it better than peers who without a coding background.
It applies to more than coding
I’ve often joked that my visual design skills usually result in something that looks like it was designed by an engineer… because it was. Boring boxes everywhere.
While AI-generated user interfaces are generally a step above that, if you have UI design skills, you have the opportunity to make a dramatic difference in the result. Again, your peers with poor or no design experience are at a disadvantage. You, on the other hand, have the know-how to instruct AI to generate something significantly more useful and usable.
This applies equally to other domain-specific skills. If you’re a game designer, you design better games than those who are not, with or without AI. If you’re a mathematician, you generate better mathematical tools, with or without AI. The list goes on.
AI is a tool, not a replacement for skill
What it all boils down to is that AI — be it used in chatbots, vibe-coding, or other tools — is just that: a tool. And like any tool, you need to know how to use it.
Using any tool effectively requires skill. Sometimes that skill is tool-specific, but sometimes those skills come from other, related areas. And when one tool is built on top of another, as vibe-coding is built on traditional coding and computer programming, those other skills pay off.
Do what you love
Twenty years ago, I wrote:
At the risk of beating the word to death, it’s all about passion.
If you find something you truly love to do, you’ll be self inspired to learn more, do more, and do it better. In fact, that passion will set up a virtuous circle of learning, doing, and improving that will continually make you more valuable at what you do every day.
If you don’t enjoy what you’re doing — if what you’re looking at doesn’t speak to you, doesn’t make you want to stay up late and get up early just so you can do it more — it might be time to reassess.
I know it’s not always practical to just pick it all up and refocus, but in the long run, work that you feel passionate about can only reward you in many, many ways.
I think that still holds true, be it about coding or anything else.
The good news is that I think that the death of coding has been severely overstated.
Do this
If you want to get good at vibe coding, I recommend starting with traditional coding fundamentals.
Will traditional coding remain a valuable skill? I believe it will be, particularly for those who are truly good at it. It still has a place, and, of course, you’ll be more adept at helping those vibe-only-coders figure out why their program isn’t doing what they expect. 
The cream rises to the top. Be the cream.
Subscribe to Confident Computing! Less frustration and more confidence, solutions, answers, and tips in your inbox every week.
I started out writing code from the specs for the project by an analyst. Then, I progressed to speaking directly with the clients who didn’t know what they needed. Later, I was the analyst giving instructions to the programmers as to what was needed. At that last point, I would be giving the same instructions to the AI chatbot and the refining process would be similar.
On my first IT job interview, I was asked, “What can you program?” I answered “I can program anything if given the specs.” He answered, “That’s the problem, you never get the specs.”
That’s what the bot can’t do, analyse the users’ needs and translate them into useable specs.