Career Advice - Stuck in the Middle Ground: Overcoming AI reliance to become a self-sufficient engineer

Career (You Can Do This!)
1

Hello everyone,

I am here to ask some question regarding my career. I have a bachelor of software engineering and currently I am doing the second year of my master’s degree in computer engineering: artificial intelligence and data analytics. In the future I would like to work in AI systems engineering. But I have a very big problem. I used AI to pretty much write the code for all of my assignments and projects until now, but I did study for exams on my own and actually graduated with high honors in my bachelor’s. But now I am in some strange middle ground where I am not a beginner (I started a python course on Exercism thinking that was my problem but I realized I knew most things they were teaching and I attempted some random exercises in codewars and actually managed to solve them. I only needed to google some small syntax specifics), but at the same time I am not that good because I cannot build something from scratch entirely by myself. I always need to consult the AI for the initial planning and for development advice, even to generate parts of the code. Because of the academic rigor of my studies, it pretty much becomes impossible to do everything without using AI to help.
My biggest problem is that it is not very clear to me what specific skills I need to have. I don’t know if I am where I need to be or if I am behind. Because I see my peers doing all kinds of projects and other things and they seem very confident and seem to know what they are talking about in tech related topics. On the other hand, I feel very distanced and I think I expected more from myself. It is honestly even kind of embarrasing that I am on the road to finish a master’s degree and still can’t build a full project on my own.

Now I am working on 2 projects related to my unversity, one of them requires me to build an AI orchestrator using LangGraph and the other to design the architecture for transitioning from an LLM to an SLM based approach. These seem very advanced to me and I think it is normal to use AI to help, but what am I actually learning if AI is telling me to do this and that?

So how can I get out of this situation? Like how is the transition from academics to the industry supposed to look like? Because most of the advice I see on the internet these days is either related to people who are self-learning or who followed their studies properly, even though my problem seems to be relatable to a lot of my peers as well. I tried asking on reddit for advice once as well and all I got was people telling me it’s my fault and to figure it out on my own.

I also want to blame the universities I attended to some extent for this. Now I am doing a masters that should be mainly focused on AI, but I have mandatory courses on computer architectures, operating systems, internet system security, networks which often have more credits than the AI courses.

So I don’t know if anyone could provide some information and advice to me because I am very confused.

Thank you in advance.

3 Likes
2

Build a project yourself using everything you know. If you don’t know something then you know where to start in your learning. It will take some to catch up but it will be worth it. Try also recalling as much as possible and try teaching someone else everything you know. If you are not able to teach it, then you know where the cracks are. It’s sort of like learning a language. If you can’t converse with someone in it fluently then you just aren’t fluent and have to keep practicing. In this case, build more using what you know and turn ai tools off.

Honestly I find myself googling the answers the old fashion way and self teaching with stack overflow. I’m not a comp sci major at all though. You can do this!

I don’t know what a comp sci education entails but you have the skills + help! You also have a network advantage so use all you can. Those advisors at your uni can help you too. Corporate coding is very different than what I do as a freelancer but if you really want to be in the industry I am sure you will find a way <3

3

for what it’s worth, a lot of early stage teams hiring for AI roles care less about whether you can write everything from memory and more about whether you can explain why the system is built the way it is. on the LangGraph one, maybe try writing the plan yourself first, even a rough one, then ask AI to poke holes in it instead of asking it for the plan. you’ll notice pretty fast which parts you actually understand

4

This is a great topic. AI can help with coding, but taking time to solve problems on your own is what builds real skills and confidence. Making mistakes and learning from them is still a big part of becoming a good developer.

2 Likes
5

To be honest freelancing is actually what I want to lean into more than regular corporate. But I need to have some job experience first.

Thank you so much for the advice as well!

6

Thank you. I will try to do that

7

No. If you can solve Codewars problems and most of Exercism, you’ve got the basics. The real difficulty is building something from scratch, and almost everyone struggles with that at this stage. The fact that you’re being honest about it is a good sign. And also you learn from what you actually understand. A simple test: close the AI and explain a part of your project in your own words. If you can’t, you haven’t learned that part yet. For the LangGraph and SLM project, you should at least be able to say why you chose this architecture, what each node does, and what happens if you remove it. :slight_smile:

1 Like
8

Start by making projects for other people–you can start with people you know. That’s how I started and that’s some of the advice in Quincy Larson’s developer job book. You have to create your own luck, and lady luck favors the one who tries. The law of serendipity

9

I have no formal education regarding this other than some HTML certification and now attacking this and JavaScript. But I saw this and thought am I missing something here by using AI in spats like only to really talk through any complete stumping points. But the loops ultimately reinforce whatever im missing but it is still a complex mental battle for me being reluctant to admit that im stuck and two I am nowhere near a mastery level . Nonetheless I appreciate your insight

10

So there’s really two things:

But now I am in some strange middle ground where I am not a beginner, but at the same time I am not that good because I cannot build something from scratch entirely by myself. I always need to consult the AI for the initial planning and for development advice, even to generate parts of the code.

This is fairly common, as once you get past some initial syntax and theory you start getting into the “main part” of programming, which is combining technologies, concepts, syntax to actually do something.

I usually use the analogy of learning to play chess. Once you learn the moves of all the pieces, and the “mechanics” of the game (the syntax), you only just start your journey into tactics, strategy and theory itself.

The fact you have become aware of this is usually a good sign, as often times learners will think syntax is all there is and get concerned that they don’t know enough to continue beyond syntax. Because like chess the knowledge beyond the syntax is essentially infinite. Unlike chess, which is a game built on infinite positions with a select number of pieces, learning is an infinite task with infinite directions! So yes its daunting, but the first step is realizing there are so many steps.

My biggest problem is that it is not very clear to me what specific skills I need to have. I don’t know if I am where I need to be or if I am behind.

I’d worry less about specific skills, and more on figuring out what you could/should learn next. This could be and probably is a huge list initially, but identifying is the first step in learning and you have to start somewhere.

But I can to point out this part:

I don’t know if I am where I need to be or if I am behind.

There’s no such thing as “behind” as everyone’s case is different, you aren’t in a race against anyone but yourself, so ahead/behind is entirely subjective to yourself. There’s too many factors to accurately position yourself against some vague “other” to get an idea of where you should be compared to where you are.

Instead, focus on where you are in regards to learning, and where you can go. Ignore what other people are doing as you don’t get the option to change what, or where you need to go same goes for them. Focus on your current skills, focus on getting better and focus on just learning more. Its a process more than anything, since its continuous and essentially forever.

This is usually done because colleges want you to graduate with a broad set of background, to help “connect the dots”. Its hard to be good at something if you hyper focus on just one topic and not well rounded, as most “connections in knowledge” are due to combining different aspects and looking at things from multiple angles and related knowledge between different fields. This is what you usually hear as “thinking outside of the box”. Can’t do that if you live in the box and just exclusively study the box continually!

I’ll point you to the direction I tell most people coming through here, college or self taught or not.

Look at jobs you’d apply to now. It might be intimidating but its the best direct indicator as to what people are looking for. You might find a few jobs fitting your skill level which is normal in today’s market. Look around for network connections or any resources in your college to see if you can get a leg up. Clubs, internships, or just knowing someone can all help. More the better.

Lastly I want to re-confirm that the market in some places is tough, there is no getting around it. But you seem like you are in a better spot than many with AI still being relevant, having AI experience ( and not complete reliance on it!) and a mix degree (engineering can lean more hardware, look for lower level jobs) you have a wide range of experiences on your college credentials alone.

Build a few projects to expand on your knowledge and look for extra curricular opportunities if possible at school and good luck!