The layoffs are coming. The robots are taking over. Better start farming mushrooms in the woods and bartering with goats.
Or, maybe… don’t panic?
AI isn’t here to steal your job. But someone using AI will—if you don’t learn how to use it yourself.
That’s not a threat. It’s just how tools work.
The Fear
I get it. You spent years building muscle memory around your editor, your framework of choice, and the sacred art of rubber duck debugging. Now some flashy autocomplete box is doing things in 0.5 seconds that used to take you an afternoon, a whiteboard, and three Slack meltdowns.
It feels insulting. Like all that craft is being replaced by a guess-and-check machine that never sleeps.
But here’s the truth: AI is a tool. The people who thrive in this new wave aren’t necessarily better programmers—they’re just better operators.
Let’s Define “Use AI”
If your only exposure to AI tools is hitting ⌘+Enter in Copilot and watching it hallucinate a sorting algorithm from 1983… yeah, that’s underwhelming.
Using AI well means knowing what to ask, when to ask, and how to filter what comes back. Here are some high-leverage ways to stay sharp:
Exploration
Part of our job is to explore solutions and come up with ideas. Maybe you'll need to evaluate a framework or service offering too. AI can be a massive help with that. For instance:
- Show me the main differences between React and Vue, and an example of pulling data from a remote API.
- Why would I choose Astro over Hugo for my blog. How about my company site?
- I'm trying to learn protobuf. Fetch this website (the protobuf docs) and explain it to me like I'm 10.
- Use this SQL file and show me what a Prisma schema looks like.
These aren’t “write code for me” prompts. They’re “give me eyes I don’t have right now.” Think of AI as the coworker you wish you had: not afraid to Google like a demon, never tired, and doesn’t judge your 2AM commits.
Refactoring
One of the best uses of AI is cleaning up the mess you already made. This is extra true if you can use customization, like "custom instructions" with Copilot.
- Turn this callback spaghetti into async/await.
- Break this 300-line React component into smaller ones.
- Replace Lodash with native functions.
- Convert this service from Dapper to Entity Framework. Yes, it’ll try.
You still need to review and test. But the energy you save not rewriting boilerplate? Use that to actually think.
Tests (Yes, Real Ones)
AI is great at writing boring tests you didn’t want to write:
- Unit tests for a service layer.
- Snapshot tests for a component you’re about to mess with.
- Load test setup scripts using your existing endpoint structure.
- Write your Gherkin BDD bits and have Copilot generate the actual specs!
Again, don’t blindly trust it. But when you’re short on time and high on ambition, it’s like having a junior dev who never complains and mostly behaves.
UX and Copy Tweaks
I use this all the time, and it's so very useful:
- “Rewrite this error message so it doesn’t sound like HAL 9000.”
- “Give me 3 alternative empty-state messages with a friendly tone.”
- “Update the onboarding copy to reflect our new feature set.”
- "Translate the i18n file into Esperanto."
Little stuff. Easy to ignore. But if you do this well, you’re the person who notices the details. That makes you very, very hard to replace.
Repetitive Tasks
Some teams spend hours every sprint:
- Creating Jira tickets.
- Writing changelogs.
- Documenting APIs.
- Wrangling Markdown tables.
Ask yourself: “Can this be templatized?” If the answer is yes, there’s probably a prompt that’ll get you 80% of the way there.
Save your creativity for the stuff that matters. Let the machine do the part where your soul usually dies.
The Shift
There was a time when “knowing Git” was optional. Same for responsive CSS. Or bundlers. Or Docker. Now? Not a chance! You need to know these tools at a minimum.
AI is heading that way. Fast. Yes, there will be people out there who generate their entire app and cackle about it on YouTube. Check back in a year and see how much they hate it. This isn't about AI doing your job for you - it's about using AI to do a better job.
You don’t need to be an expert. You just need to show that you’re adaptable. That you’re not waiting for someone else to tell you what’s safe to learn.
One More Thing
If you’re worried about AI taking your job, here’s the question to ask:
What would I automate if I were my boss?
Then… do that. You’re not here to fight the machine. You’re here to drive it. And yeah, it might crash sometimes. But at least your hands are on the wheel.