I don't have a solution to offer you. I wish I did. But I do have a warning.
If you're trying to break into software engineering right now, in 202...
For further actions, you may consider blocking this person and/or reporting abuse
his is a powerful and honest reflection on what breaking into software engineering feels like in the AI era.
As someone coming from a mathematics and data-analysis background and currently building my coding skills step by step, I feel this “20-foot burning gate” reality very deeply.
Instead of chasing surface-level tutorials, I’m trying to follow the path you described:
– building real small projects
– focusing on debugging and understanding code, not just generating it
– using AI as a learning partner, not a shortcut
I’ve recently started sharing datasets and notebooks on Kaggle as part of this journey.
If you have time, I would truly value your feedback on my work and your honest opinion about my chances as a new coder trying to enter the field in 2026.
Your article doesn’t just warn people — it gives direction.
Thank you for writing something this real.
thank you for engaging in this piece.
coming from a math and data-analysis background is a pretty strong foundation right now - you're approaching things, small and real projects, prioritizing understanding and debugging, using AI as a learning partner is a good thing here.
keep going, the gate is high, but focusing on judgment over output puts you in the group that still has a real shot!
Thank you, Adam — I truly appreciate your thoughtful response and encouragement.
Coming from a mathematics and data-analysis background, I’ve always believed that understanding, judgment, and the ability to debug reality matter more than simply producing code output.
Your article resonated with me because it describes the landscape honestly, without removing hope from the people still willing to build, learn, and think deeply.
I’m trying to focus on small but real projects, using AI as a learning partner rather than a shortcut, and gradually strengthening the kind of decision-making that software and data work really require.
Hearing that this direction still gives someone “a real shot” means a lot.
Thank you again for taking the time to engage — and for writing something that pushes many of us to aim higher rather than give up.
I would say you're probably in a good place, because you've got specialized expertise - not "just" generic web dev skills, but math/statistics/data analysis - I think that puts you in a completely different position ...
It would be like a coder/developer who also possesses domain knowledge in another field (accounting, medicine, you name it) - they would be able to work in that field/industry and have an obvious edge over a developer without that domain knowledge ...
Emphasize your math and data analysis skills, and you should be in a strong position when you apply to jobs requiring those skills.
Thank you, I really appreciate this perspective — it highlights something I’ve been reflecting on a lot.
Your point about domain knowledge creating a real edge resonates strongly with me. Mathematics, statistics, and data analysis don’t just add extra skills; they shape the way problems are understood, structured, and solved. That deeper analytical lens feels increasingly important, especially in a time when generic coding output is becoming easier to automate.
I’m trying to lean into that strength by building projects that emphasize interpretation, clarity of insight, and real analytical usefulness, rather than just technical implementation. Hearing that this direction can translate into a meaningful professional advantage is genuinely encouraging.
If you ever have time, I’d truly value your thoughts on a couple of my Kaggle works — particularly the Employee dataset and the Swiss Army Knife dashboard notebook on Kaggle — since both are attempts to express that blend of analytical thinking and practical tooling. Any honest feedback or guidance would be greatly appreciated.
Thanks again for sharing such a thoughtful and motivating insight — conversations like this make the path forward feel much clearer.
You've got a genuine advantage by having "domain knowledge" in what I would consider a different (although related) field - compare it with an economist/accountant, or a biologist/geneticist, etc, who also knows how to code or develop - it might be a way out of the dilemmas which the author of this article sketched ...
(if AI continues to mature and develop in the way it does then I see a decreasing demand for "pure" hard-core developers anyway, including senior ones - probably you'll get more people who do development as part of their job in a different or adjacent field - rise of the "citizen developer" as I believe they call it)
Absolutely — that comparison makes a lot of sense, and it really reframes how I’m thinking about development in the AI era.
Having deep domain knowledge in math and data analysis feels like a genuine strategic advantage, much like a biologist who codes or an accountant who automates financial workflows. It allows you to approach problems from a perspective that pure coding skills alone can’t provide, and it creates opportunities to contribute practically valuable solutions rather than just code output.
I also see the point about the evolving role of developers. As AI matures, I agree that the landscape will favor people who combine coding with domain expertise — the “citizen developer” model seems likely to grow. That’s exactly why I’ve been experimenting with projects like my Swiss Army Knife dashboard and the Employee dataset: they’re attempts to integrate analysis, domain insight, and usable tools in a single workflow.
I’d genuinely value your take on them if you have a chance — any feedback on usefulness, structure, or ways to make them more practical would be really helpful for guiding my next steps.
It’s exciting to think about how domain-driven coding could be one of the ways forward in this AI-heavy landscape.
Yeah sure no problem - how to access them? I don't have Kaggle installed, can you enlighten me?
Absolutely! You don’t need to install Kaggle — both projects are fully accessible via web browser. Here are the direct links:
Swiss Army Knife Dashboard: kaggle.com/code/ahmedanterelsayed/...
EmployeeDataset: kaggle.com/datasets/ahmedanterelsa...
You can view all notebooks, datasets, and outputs online — no download required. I’d really value any feedback you might have on them!
There is one thing you overlooked. People with money to spare can build applications because they can spend money on AI.
And there is why I agree with you software development isn't a meritocracy anymore
right. access to AI isn't free. those who can afford paid models, better tooling, cloud resources and the time to experiment are starting from a completely different position than those who can't.
when " build impressive projects " becomes the baseline expectation, we're basically just saying " you need capital to even compete "
This is a brutal reality check. As someone just starting college, it feels like watching the bridge I’m supposed to cross burn down while I’m still on the bank.
The 'Middle Ground' is a terrifying place to be right now. I haven't reached my potential yet, but I’m already being asked to have Senior-level judgment just to get an internship. It feels like the goalposts aren't just moving; they're being automated.
However, maybe starting now is an advantage? I don't have to 'unlearn' the old ways. I can learn Forensic Debugging and Systems Design as my first language. If the 'coder' is a commodity, I have to become a Product Architect who uses AI to build things that were once impossible for one person.
The gate is definitely on fire, but I’d rather know that now than find out after four years of tuition.
Yeah a bit doom & gloom, but obviously that's the world we live in, change is the only constant ...
TBH the path proposed sounds like it might not be worth it for all of the people who would previously have tried to get into software dev - I mean, is this still worth it? I don't envy newcomers who have to tread this path ...
A way around the problem might be to start your own company (business), or just choose a career in another field - plumbers, farmers, doctors haven't been AI-ized yet - or, still aspire to become a developer, but also have domain knowledge in another field (hey, that "other field" could even be AI ?)
P.S. another point - I think there's also a role to play for universities and other educational institutes to adapt to these new realities - if companies or employers ask for different skills, then they should try to evolve their programs to reflect that ...
I'm not saying you're wrong, but I do say you might be a bit too pessimistic - I have an antidote for you, read this article, and then especially the sections "The Productivity Paradox" and "Where the Bottleneck Actually Is":
dev.to/sudheer_singh_3329d404bb1/o...
and I commented on it as well:
dev.to/leob/comment/34cn1
But, I admit that this might not change the way companies are looking at it (at least in the short term), and might not stop them from thinking they can save a quick buck by getting rid of their juniors (or by not hiring any) - because the short term apparent productivity gains might look obvious, and the hidden costs might now ...
But maybe they'll find out at some point.