Should You Leave Software Engineering? A Calm Way to Decide

You've googled "should I leave software engineering" more than once, and probably late, and probably right after a layoff announcement or a model demo or a LinkedIn post from someone declaring the profession finished. The tab is open next to a half-drafted message to a friend who "got out" and now does something that sounds safer. The impulse is real and it's loud: get off this sinking ship before it takes you down.

Here's what's worth noticing before you act on it. You are asking one of the biggest decisions of your working life at the exact moment you are least equipped to make it, flooded, sleep-deprived, and reasoning from a threat you can't size. That's not a character flaw; it's how spirals work. But you would never approve a production migration in this state, and this is bigger than a migration. The move is not to answer the question right now. The move is to refuse to answer it from a panic, and to build the conditions where a calm version of you can answer it instead.

Panic is a bad architect

A decision made in a spiral has a signature, and you can spot it. It's binary (stay and die, or leave and be saved). It's urgent (must decide this week). And it's driven by a threat you can't actually quantify, which is exactly what makes the urgency feel unbearable. Compare that to how you'd approach any hard technical decision at work: you'd gather constraints, weigh trade-offs, sanity-check assumptions, and explicitly resist the pressure to commit before you understand the problem. The career version deserves the same discipline, and the first act of discipline is separating the feeling (I need to escape) from the decision (is leaving engineering the right long-term move for me).

The feeling is often not even about engineering. AI dread almost never travels alone. In the forums where developers talk honestly, the "should I leave" post is reliably stacked on top of something older: imposter syndrome that predates the models, a new baby, a mortgage, family obligations, a stagnant role, plain work stress that had nothing to do with AI until AI gave it a dramatic new name. If you leave the field but carry the fused bundle with you, you've relocated the problem, not solved it. So before the career question, an honest diagnostic: how much of this urge is "engineering is wrong for me" and how much is "my life is heavy right now and this is the thread I can see to pull"?

Runway first, then the honest inventory

The people who navigate this well, across hundreds of forum threads, cite the same sequence, and the first step is not soul-searching, it's money. Build runway before you decide anything. A financial buffer doesn't just pay bills; it changes the kind of decision you're making. With three weeks of savings, "should I leave?" is a panic question with a gun to its head. With a year of expenses banked, it becomes a calm strategic question you can actually think about, because the catastrophe timeline is gone. This is why "live well below your means in a volatile industry" is the most upvoted career advice in the field, not because frugality is virtuous, but because runway converts an existential dread into a solvable problem with a timeline. You cannot decide clearly while you're falling; step one is stop the fall.

Then, and only then, run the honest inventory, and make it specific. Not "am I good at coding" but a real audit of where your value actually lives. Split your work into two columns. One is syntax: the stuff that's the typing, the boilerplate, the lookup-able mechanics, the part the tools are genuinely eating. The other is judgment: knowing what to build and what to refuse, reading a vague requirement and finding the real need under it, debugging a system nobody understands anymore, deciding what's safe to ship on a deadline, translating between the humans and the machine. The forums are blunt about this: your job was never typing code. If your honest inventory says most of your value was in column one, that's genuinely useful information, but the answer is probably "grow the judgment half," not "flee the field." If it says you've been doing column-two work for years without noticing, you're far less replaceable than the panic claims (which is the whole premise of how an engineer's value actually holds up), and the ambient dread narrative has been lying to you about your own resume.

Fleeing to "safe" is a trap with two anxieties

Here's the failure mode to name out loud, because it's the one the spiral is steering you toward. You quit for a field that sounds safe from AI, chosen at the bottom of a spiral, on the theory that anything is better than the countdown. And you often discover two things at once: the "safe" field has its own disruptions, its own precarity, and its own version of the fear you were running from, and you've now voluntarily walked away from a decade of hard-won, well-compensated expertise to start over near the bottom of an unfamiliar ladder. You didn't trade one anxiety for zero. You traded one anxiety for two: the new field's instability plus the grief and financial hit of abandoning a craft you were actually good at. Panic promised escape and delivered a downgrade.

That's not an argument to never leave. Some people genuinely should, because they've lost the love, or found something that fits them better, or the honest inventory really did come back thin. It's an argument that leaving well looks nothing like fleeing. Leaving well happens from a calm mind, on top of a runway, after an honest inventory, toward something specific you're moving to, not away from a threat you never sized. If the fear itself has become the disease (insomnia, panic attacks, a dread that doesn't lift when you close the laptop), that's worth professional support before it's worth a career decision, because no life choice made from a nervous system in alarm is trustworthy. And if you're ever in a place where you're thinking about harming yourself, call or text 988. The career can wait. You reasoning clearly cannot.

Frequently asked questions

Should I quit software engineering because of AI? Not on the timeline your panic wants. Decide from a plan, not a spiral: build runway first so the catastrophe clock stops, then run an honest inventory of judgment versus syntax. If you still choose to leave from that calm place, toward something specific, that can be a great decision. Leaving from a threat you can't size rarely is.

How do I know if I actually want to leave or I'm just scared? Separate the feeling from the decision. "I need to escape this feeling" and "engineering is wrong for my next decade" are different statements that a spiral fuses together. A useful test: imagine you had two years of savings and slept eight hours last night. Does the urge to leave survive that? If it evaporates, it was fear, not a verdict.

Is switching to a non-tech career actually safer from AI? Rarely as safe as it looks from inside the spiral, and it carries a hidden cost: you abandon a decade of expensive expertise to restart near the bottom elsewhere. Most "AI-safe" fields have their own precarity. You can end up carrying two anxieties instead of zero. Choose a destination for what it offers, not for what it lets you flee.

What if the anxiety about my career is making me unable to function? Then the anxiety is the problem to solve first, before any career move. Insomnia, panic attacks, or an inability to work aren't inputs to a career decision, they're signs to get professional support, because decisions made from an alarmed nervous system aren't reliable. If you're having thoughts of self-harm, call or text 988 right now.

Deciding well requires a mind that isn't on fire. ILTY is an honest AI companion that helps engineers think from a plan instead of a panic, it pushes back on the spiral instead of feeding it. See how it fits an engineer's head, or try ILTY free.