I'm an anxious person. I've felt an extra layer of unease towards software engineering recently, and for a while I couldn't put a finger on why. I figured it was just the general newness and uncertainty of everything happening in the industry, and I pushed it away. I work as a full stack developer in the public sector, and I feel there's an extremely low chance I could be affected by the layoffs I see all around me. My team is actively working to stay current on the AI front, and I have more side projects than is likely healthy for me. So why do I constantly feel like I am being left behind?
AI capabilities and technologies are just moving so fast. I'm sure I'm far from the only person feeling this way, and I finally was able to articulate the state of software engineering from the eyes of someone relatively early in their career. I believe that all software engineers (especially young professionals) that are actively using AI will relate in some fashion to these two categories of anxiety:
- Uncertainty towards the future of the industry
- The feeling that you could be doing more
This is an article about becoming aware of the anxieties, not necessarily debating them or fixing them. But hey, awareness is half the solution.
Uncertainty Towards the Future
Let me just lay out what the last couple years have looked like for the tech industry, because it's hard to talk about the anxiety without the context. Over 150,000 tech jobs were cut in 2024. In 2025 it was even worse, north of 172,000 positions eliminated.1 Since 2022, the tech sector has shed roughly 666,000 employees across something like 3,500 companies.2 That's not a blip. That's a pattern.
And it's not evenly distributed. Entry-level positions have been hit the hardest. Entry-level tech jobs dropped 60% between 2022 and 2024.3 Software developer employment overall fell 27.5% between 2023 and 2025. A Harvard study that tracked 62 million workers found that junior employment at companies actively adopting AI declined by 9 to 10% within six quarters of implementation, while senior employment stayed basically the same.4 The math is pretty straightforward from the company's perspective: one senior engineer with good AI tooling can now do the work that used to require a few junior engineers.
Then you have the executive class making statements that feel almost designed to give you a panic attack. Dario Amodei, the CEO of Anthropic, said in May 2025 that AI could eliminate up to 50% of all entry-level white-collar jobs within one to five years. His actual words were that entry-level positions for people under 30 "are going to be eviscerated."5 Not exactly reassuring. Sam Altman said in February 2026 that "it'll be very hard to outwork a GPU."6 Cool, thanks Sam.
Shopify's CEO, Tobi Lütke, put out a company-wide memo in April 2025 telling managers that before they can request a new hire, they have to prove AI can't do the job first. AI usage is now a "baseline expectation" at Shopify and it factors into performance reviews.7 Klarna went even further, announcing they were replacing Salesforce and Workday with internally built AI systems, and cut their workforce from 7,000 to under 3,000 employees.8
I think the thing that gets under my skin the most is the narrative that Big Tech is trying to push. The story they're telling is that SaaS is dead, traditional software roles are shrinking, and if you're not on the AI train you're already behind. And some of that is legitimate. But some of it is also just CEOs justifying layoffs and cost-cutting to their boards. Dario Amodei himself acknowledged that many CEOs privately view AI as a cost-cutting tool more than a productivity enhancer.9 The narrative serves them.
Jensen Huang from Nvidia has a framing that I think about a lot. He said the threat isn't AI itself, it's "the person who uses AI to automate that 20% is going to take your job."10 Which is somehow more stressful than just being replaced by a machine, because now I'm competing against other people who might be using these tools better than me. It's not a robot coming for my job. It's a developer who figured out MCPs and subagents before I did.
I'll say this though. I work in the public sector. My job feels secure. Most senior engineers are doing fine. A lot of these predictions haven't fully materialized yet. And the people making the boldest claims are usually the ones selling the AI. I'm not trying to dismiss the data because the numbers are real, but the doom narrative has a bias to it that's worth recognizing. Still, it's hard to feel settled when every week there's a new headline about another round of cuts or another CEO saying your role is temporary.
Should I Be Doing More?
This one is more personal and probably harder to talk about. The first category of anxiety is about the industry. This one is about me.
It is so easy to hop on X and see 10 posts from 10 different people who just one-shot prompted an incredible product that's already making money. Or see talented engineers sharing their complicated automations and methodologies towards a new feature that I haven't even tried yet. That was me with MCPs and subagents. I kept seeing people build these elaborate agent workflows and I hadn't even set one up. It felt like everyone was operating on a different level and I was stuck reading about it instead of doing it.
There's this constant background hum that I am being outskilled and that if I ever needed to find a different job, I wouldn't be able to compete. And I know developers who work a full day in the backlog for their company, then come home and code on side projects just to keep their skills sharp. I do that too. But even that doesn't feel like enough sometimes.
Because the tools have made everything so accessible, the expectation of output has gone through the roof. When I can log into Codex on my phone and run a few prompts, scaffold out a feature, and create a pull request at 3am because I had an idea, how can I justify going back to sleep without getting it out of my head? I should always be shipping. I should always be learning. If I don't, I'm never going to move up in the world. At least, that's easy to believe.
There's actually data that backs this up. A recent Atlassian study found that 99% of developers report time savings from AI tools, with 68% saving over 10 hours a week. But burnout is getting worse, not better.11 That tracks. The tools got faster so the expectations got higher. You save 10 hours and your company doesn't give you 10 hours of breathing room. You get 10 hours of additional output expected. 63% of developers now say their leadership doesn't understand their pain points, up from 44% just a year before.12 The tools improved but the job didn't get easier.
But then how do I manage all of that alongside my actual job, my family, my friends, my hobbies, my ecclesiastical duties? How do I reconcile the pressure to always be learning and shipping with the reality that I'm a person who needs sleep and has a life outside of code? It's easy to get wrapped up in that line of thinking, and I catch myself in it more often than I'd like.
So What Now?
I don't have a clean answer for any of this. I'm not going to pretend I've figured it out. But I will say that writing it down helped me see it more clearly. These are two different anxieties that feed each other. The uncertainty about the industry makes me feel like I need to do more, and the pressure to do more makes the uncertainty feel heavier. Recognizing that loop was useful for me.
My biggest reassurance is that anyone who has this type of thought process, anyone who is worried about being outpaced or left behind, is probably a software engineer who actually cares about what they do, what they ship, and what they learn. The people who don't care aren't losing sleep over this. If you're reading this and nodding along, it's because you honestly care about your craft. That counts for a lot.
I have a whole career to learn these things and it doesn't have to happen overnight.
Thanks for reading.