Skip to main content

Earning after the advent of AI, the 3 paths and things that matter

I have received this question more times than I can count. Are jobs gone? Is AI taking over everything? Why are internships not coming through? What do I do with my degree or my skill?

Here is my honest answer — not a motivational one.

AI has made things easier, and harder at the same time. Because AI can now do multiple things, the person who knows how to get the right things done from AI is preferred over the one who does not. That is the new baseline. Learn the basics of AI regardless of which path you choose — not to become an AI engineer, but to use it well in whatever you are doing.

Beyond that, I see three paths. Each one has its own direction. The mistake most people make is not choosing one and going deep on it.


Path 1: Build a Personal Product

This is the path I am on. I will be honest about where my experience ends.

Start by finding a problem — ideally one you personally face. That is your built-in validation. If you have the problem, others likely do too. Finalise the idea, confirm the problem is real, then learn the skill you need to build the solution. Do not learn everything first. Learn what you need, build, then learn more.

Once the product is built, the work shifts to marketing. This is where most builders stop — because it feels like a different skill entirely. It is. But it is learnable the same way development was. Start with early users, gather feedback, improve, then do more marketing.

I am at the marketing stage myself. I cannot tell you what comes after from personal experience — only that the direction is clear and the work continues.

How to share: Talk about the problems your customer faces. Occasionally present your product as one example of the solution — not the pitch, just the example. If you run ads, use them to push that same message to more people. Not to convince anyone to buy. Just to reach more of the right people with something honest and useful.


Path 2: Get a Job

The approach here is the same one that got someone I spoke to recently his first client — focus on one thing.

Learn the basics of your field first. In AI for example: machine learning, deep learning, RAG, automation. Get a working understanding of each. Then pick one and go deep. Not because the others do not matter, but because depth in one thing is what makes you visible and hireable. A RAG engineer who knows RAG well is more useful to an employer than someone who knows five things at a surface level.

The projects you build matter too. Do not build dummy projects. Build things that solve real problems — the kind of problems an employer is already solving for his clients. That is what shows you can do the actual work.

How to share: Document what you are learning. Share projects as examples of solutions to real problems — not as a showcase of your effort. The employer coming across your work should see someone who thinks in terms of problems and solutions, not someone listing what they built. I have not done this myself so I cannot speak from experience — but the direction is clear.


Path 3: Freelancing

The learning direction is similar to the job path. Pick a niche, go deep, build relevant projects. The difference is in how you show up publicly.

As a freelancer you are not trying to impress an employer — you are trying to attract a client. That client is not looking for someone learning. They are looking for someone who can solve their problem right now. So the sharing shifts: less about what you are learning, more about what you are doing, what you are working on, how it helps, why it matters.

Same angle as everything else — lead with the problem, present your work as one of the solutions. The client who reaches out already understands what you do and why it is relevant to them.


The thing all three paths share

In every path, the way to get clients or employers is the same: inbound, not outreach.

Outreach costs time and energy. Cold messages ask a stranger to trust you before they have any reason to. Inbound builds that trust before the conversation starts — through consistent, specific, honest work shared publicly over time.

Answer the questions your audience is already asking. Talk about what they are thinking. Share your work in a way that shows you understand their problem. The right people find you.

This is the core of They Ask You Answer and Show Your Work — two books worth reading if you want to understand the implementation deeply, not just the idea.


A few things that apply regardless of path

Keep learning continuously. The field moves and stopping is not an option.

Set a clear goal. Pick a path and know what you are working toward. Vague direction produces vague results.

Stay consistent when the phase changes. Development to marketing. Learning to sharing. Building to selling. Each transition feels like starting over. It is not. It is just the next stage of the same work. Most people quit here. Do not.

And learn AI basics regardless of your path. Not to become an AI specialist — to use it well in whatever you are already doing. The person using AI effectively will always have an edge over the one who is not.


Jobs have not disappeared. The bar has shifted. The people getting hired and getting clients are the ones who picked a direction, went deep, and made theirselves visible to the right people, the right way.

This is the whole answer.

Comments

Popular posts from this blog

A user found the bug I should have caught myself

I was collecting user feedback on Iqra a few weeks ago. One of the users pointed out a bug. A section of the app was not working correctly. I knew the section. I had made changes to it recently — and shipped them directly to production without running through my usual testing flow. A user told me about the bug. Not my emulator. Not my test release. A user. I fixed the bug, made a few more improvements, added some things to the app. When I finally sat down to build the test release properly, this came back to me. So here it is — written down, for anyone building a product right now. The flow that actually protects you Every change, no matter how small, goes through three stages before it reaches your users. First, test on the emulator. Catch the obvious breaks early, before anything leaves your machine. Second, publish a test release and run it on a real device. The emulator lies sometimes — screen sizes behave differently, performance feels different, interactions that ...

How an AI Student is Earning from LinkedIn — and What I Learned from a 17-Minute Call

A few days ago I saw a post on Facebook. Someone sharing that he had started learning AI, done some projects, and was now earning from it. I asked him how. He said he got a client through LinkedIn. I asked him to elaborate. He said let's talk. Seventeen minutes on WhatsApp. That was the whole conversation. But it connected more dots than months of reading about "the opportunity in AI." The problem with how most people learn AI AI is a wide field. Machine learning. Deep learning. Model development. Model deployment. Data analysis. Data analytics. Automation. Most people learning AI are aware of all of these — and that awareness becomes the problem. They keep exploring, keep switching, keep adding things to the list. A year passes and they are still learning. He told me something simple: to stand out and to get clients, you have to focus on one thing. His one thing is RAG — Retrieval Augmented Generation. Not AI in general. Not even a broad category within AI...

Bhai aap ny Flutter kese sikha

Bhai aap ny Flutter kese sikha — yeh sawaal mujhse kai baar pucha gaya hai. Aaj honestly bata deta hun. Seedhi baat karta hun. Maine Flutter sikhne ke liye koi special course nahi kiya. Koi bootcamp nahi. Koi structured plan nahi tha. Mere paas already C++, C#, HTML, CSS, JavaScript, PHP, Python ka background tha. Toh Flutter ek nayi language ki tarah nahi thi — ek nayi duniya thi jisme samajhna tha ke cheezein kaise chalti hain. Maine 2-3 full-length videos dekhe. Har ek 1 hour se zyada ki. Lekin goal yeh nahi tha ke sab yaad ho jaye. Goal sirf yeh tha ke flow samajh aaye. Components kya hain, cheezein kaise connect hoti hain. Bas. Phir maine project shuru kar diya. Aur yahan se asli story shuru hoti hai. Pehla hafte — setup. Sirf setup. Dependencies download hoti thi. Ghanton tak. Mere paas 4Mbps connection hai — ek typical Pakistani developer ki tarah - shayd kch zyada. Ek cheez download hoti, kuch aur missing hota, phir kch aur, phir kch aur. 7 din. S...