If you start in one of six specific jobs today, there’s a high probability that in a year, you’ll struggle to find work. This isn’t a reflection of your skills, but a testament to a rapidly changing market. In this article, I’ll show you how to avoid this trap.
Before we dive into the list, you must understand a fundamental economic principle that has held true for centuries: any job with a 100% predictable outcome is a job for a machine, not a human.
Artificial intelligence, particularly Large Language Models (LLMs), is essentially a prediction tool. AI doesn’t “understand” in the human sense. It excels at accurately predicting the correct output for repetitive tasks. If your work is all about repetition and producing the same result every time, you are in direct competition with a machine built for that exact purpose.
The market no longer wants someone who just does. It wants someone who thinks. It doesn’t need a doer; it needs a planner.
Let’s get into the list. While the first couple of roles are tech-focused, the rest apply to a much broader audience.
1. The Syntax Translator (The “Code Monkey”)
There used to be a role, often called a “code monkey,” for developers who performed simple, repetitive tasks. “Take this template and convert it to HTML and CSS.” “Set up this boilerplate.” “Spin up a few basic CRUD endpoints.”
These tasks are incredibly simple and highly repetitive. AI can now do them in seconds.
The Solution: You must start seeing the bigger picture. Deep specialization is no longer the safest bet. It’s far better to understand the entire Software Development Life Cycle (SDLC). Become a full-stack developer. I’ve said this many times: a single person with a broad understanding, aided by AI, can outperform two highly specialized individuals.
You need to be that one person who grasps the entire system. What is system design, really? You have to start thinking like an architect.
“But I’m just a junior, I just graduated, I’m still learning!”
That mindset is obsolete. You cannot compare yourself to someone who was in your position five, let alone eight, years ago. The landscape has completely changed. You have tools at your disposal that previous generations of developers couldn’t even dream of. The “junior” label of the past doesn’t apply to the new generation of talent.
You must also understand the project from the product perspective. This was always a fundamental principle, but deep specialization caused developers to lose sight of the “why” behind their work. You need to reconnect with the product as a whole.
2. The Manual QA Tester
Not long ago, companies hired armies of manual testers. I remember working on a project where we had more testers than programmers. It was a massive migration of over a thousand marketing websites for car dealerships. While the core code was written once, each site had customizations that required individual testing.
Today, this is a simple task for AI. It can generate countless bug scenarios, write scripts to click and interact with a site, and analyze the results. AI agents can now simulate a real user, behaving and thinking like a human would.
Does this mean the testing field is dead? Of course not. Just like programming, it has evolved. The competition is fiercer because the demand for purely manual work will decrease.
The Solution: You must learn to use AI in your workflow. Be the one who uses AI to replace those who don’t. Master the fundamentals of test automation. While any good QA professional can write test cases, you can now use AI to do it at lightning speed, freeing you up for higher-level tasks.
The real game-changer? Use AI to discover edge cases. The most valuable QA professionals are the ones who find the bizarre, unexpected bugs. Think of the classic meme: “A QA engineer walks into a bar. Orders a beer. Orders 999 beers. Orders a lizard. Orders -1 beers. Orders a ‘sfgshdf’.” Use AI to brainstorm these strange scenarios and elevate your value.
3. The Content & SEO Writer
This category includes content creation for SEO, general writing, and translation.
AI models have become frighteningly accurate at translation. You see it on social media—one click translates content from any language, sometimes even dialects, with remarkable precision. This will only improve as models are fine-tuned on specific languages and dialects, which is a trivial task for an LLM. We’re already seeing dedicated translation tools emerge that will rapidly mature.
The same applies to content writing. You might argue that AI-generated content feels soulless and robotic. But no serious writer today works without AI. They may not let it write the entire piece from scratch, but the time it takes to draft an article has been drastically reduced. Anyone with an idea can now prompt an AI, get a draft, and refine it into a strong piece of content.
Furthermore, the world of SEO is in upheaval. Search is not what it used to be. Many are finding that even paid Google Ads aren’t delivering the returns they once did, as users shift towards video and other forms of media.
The Solution: Become a Subject Matter Expert (SME). When you ask an AI to write about a highly specialized topic—like medicine, law, or financial analysis—it is very likely to “hallucinate” and produce incorrect information. This is where you come in. Your expertise allows you to guide the AI and correct its mistakes. While one could argue that a model could be fine-tuned on a specific business domain to reduce hallucinations, the need for a human expert to verify and refine the output remains crucial.
4. The Basic Graphic Designer
I’m talking about the designer who works in Photoshop creating simple assets. Consider thumbnails for articles or social media. Whether a thumbnail is “good” is subjective, but the ultimate measure is performance—does it get clicks? I’ve received countless messages from designers offering to create better thumbnails for me, often showing examples that were clearly 100% AI-generated.
My first thought is always: why would I pay you when I can write the same prompt myself?
AI models for image generation and photo editing are advancing at an incredible pace. The role of the basic graphic designer is in jeopardy.
However, this doesn’t apply to everything. Motion graphics is a different story. This field is more complex and, for now, safer. While I’ve experimented with models that can generate simple motion graphics from a text prompt, creating high-quality, production-ready animations is still incredibly difficult for AI. We’ve seen AI companies invest heavily in video generation with underwhelming results. It seems LLMs have hit a limit in the video domain, and motion graphics is an even more niche and complex area. This field is not facing immediate replacement.
5. The Video Editor
There’s a difference between video editing (cuts, zooms, transitions) and motion graphics (animated text and illustrations). I’m currently working with an AI-powered editor that operates on the same principles as AI-assisted coding.
You simply talk to the AI. “Make a cut here.” “Add a zoom effect on this phrase.” “Analyze this section and suggest an effect.” You work without ever touching the timeline. The cuts, effects, and transitions are applied automatically.
I speak from experience as someone who used to spend 15 hours editing a single article to achieve a certain quality. With AI assistance, I can now achieve the same quality in under two hours.
Will this replace all video editors? No. An inexperienced person telling an AI to “edit a video” will get bizarre results because they don’t know how to direct the AI. It’s the same as with programming; you need to be a good programmer to use AI coding tools effectively. The same holds true for video. This job will see a dramatic increase in competition as the tools become more accessible, but expertise will still command a premium.
6. The Voice: Call Centers & Voice-Overs
This category includes any job that is purely about voice output, with no creative or critical thinking required. Call center agents, voice-over artists for simple narrations, and similar roles are at extreme risk.
In these tasks, hallucination is almost non-existent. The job is deterministic. If the output is bad, you just run it again. The task is pure repetition, and AI models for audio are already magnificent. They can handle any language, any dialect, any accent. We are already seeing customer support being handled by AI, and as people get used to it, it will become the norm. The role of the human voice for repetitive tasks is fading fast.
The Way Forward: How to Win
Does this all sound pessimistic? If you’ve listened closely, you’ll realize the message isn’t that the world is ending, but that the competition has become fiercer.
So, the difficult question is: what makes you special enough to win?
The demand for building software and integrating AI is exploding. You must become fluent in the tools of AI, and not just by chatting with a generic chatbot. You need to understand how these systems work and how to leverage them on a larger scale. Immerse yourself in AI.
Crucially, you must learn how to market yourself. How do you build a personal brand? How do you stand out in a sea of noise? In our era, attention is the most valuable currency. You have to find an authentic way to capture it that works for you.
Don’t despair because of the intense competition. That is the nature of the world. Survival has always been for the fittest. I want you to be the fittest. I want you to be the one who wins.
Most people do not understand what is happening. They are like the proverbial frog in boiling water, feeling the heat rise but never jumping out because they don’t know what lies beyond. Do not be that frog.
Let me remind you again:
- Move away from repetition. Move towards decisions.
- Be the person who uses AI to replace someone else, not the one being replaced. Be the best at using AI.
That idea you’ve been dreaming of? That ambition you have? Start now. Continue working on it. Don’t waste another moment. You never know what might succeed.
Many people will give up. Please, do not be one of them.