惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

罗磊的独立博客
Google DeepMind News
Google DeepMind News
MyScale Blog
MyScale Blog
A
About on SuperTechFans
Martin Fowler
Martin Fowler
M
MIT News - Artificial intelligence
Recent Announcements
Recent Announcements
D
DataBreaches.Net
B
Blog
博客园 - 【当耐特】
爱范儿
爱范儿
有赞技术团队
有赞技术团队
P
Proofpoint News Feed
WordPress大学
WordPress大学
小众软件
小众软件
Apple Machine Learning Research
Apple Machine Learning Research
I
InfoQ
Engineering at Meta
Engineering at Meta
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
Last Week in AI
Last Week in AI
Microsoft Azure Blog
Microsoft Azure Blog
雷峰网
雷峰网
量子位
G
Google Developers Blog

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
Am I a Developer or Just a Prompt Engineer?
Harsh · 2026-05-05 · via DEV Community

Three years ago, if you asked me "what do you do?" I had an answer I'm a software developer. I write code. I fix bugs. I solve problems.

Confident. Clear. No hesitation.

Last week, a junior developer asked me the same question What do you actually do?

I opened my mouth. Nothing came out Not because I forgot. Because I genuinely didn't know anymore I write code, I finally said. "But AI writes most of it."

So you're a prompt engineer? they asked.

I laughed. Then I stopped. Because the question wasn't wrong Three years ago, I knew who I was. Today, I'm not sure.

This isn't an anti-AI article. It's not about going back. It's about waking up one day and realizing you don't know what to call yourself anymore.

Am I still a developer? Or did I trade the craft for a faster way to ship?


What I Used to Say

A few years ago, if someone asked what I did, the answer came easily.

"I'm a developer. I build software. I solve problems with code."

That answer had weight. It described not just what I did but who I was. There was something solid in it something that felt earned.

I'd spend weekends on side projects nobody asked for. I'd refactor the same function three times not because it needed it, but because making it elegant was its own reward. I'd debug for hours, not because it was the efficient choice, but because finding the bug felt like winning something. A small private lottery that only I knew I'd entered.

The code was mine. The struggle was mine. The satisfaction was mine.

I'd read other people's code just to see how they thought. I'd have opinions about architecture. Strong ones. I'd argue about naming conventions longer than was reasonable, because the names mattered to me, because the code mattered, because I was in it.

That person feels like a different person now.


The Shift I Didn't Notice

It didn't happen overnight. That's what makes it hard to point to.

First, I used AI for boilerplate. The tedious stuff scaffolding, config files, the repetitive patterns I'd written a hundred times. No identity loss there. Smart move.

Then, I used it for functions I could write but didn't want to. Faster. Still felt fine.

Then, I used it for functions I should have known. This is where I should have paused. I didn't.

Then, I stopped writing code first. I started prompting first. Why struggle with something for twenty minutes when AI can produce a working version in ten seconds?

Then, I stopped evaluating the output carefully. I started skimming it. Shipping it.

Then, last week, a junior developer asked me "what do you actually do?" — and I had nothing.

The shift wasn't a decision I made. It was a thousand small yeses, each one feeling like efficiency, none of them feeling like losing something — until I looked back and couldn't find the person I used to be.

That's the thing about gradual loss. You don't feel it happening. You only notice it's gone.


So... What Am I Now?

A prompt engineer writes prompts. A developer builds systems.

I still do both. I still think about architecture. I still care about edge cases. I still debug though less often, and less deeply than I used to. I still have opinions about how things should be built.

But I also spend a significant part of my day generating, skimming, accepting, and shipping code I didn't fully think through. Code that works. Code that isn't really mine.

So where's the line?

Here's the honest answer I've landed on, after weeks of not wanting to say it out loud: I'm both. And neither. And the ratio is what actually matters.

I'm a developer when I'm designing the system when I'm reasoning about trade-offs, when I'm catching what the AI missed, when I'm asking "is this the right solution" instead of just "does this work."

I'm a prompt engineer when I'm just generating and shipping. When I've outsourced not just the typing, but the thinking.

The title doesn't matter. The ratio does.

Am I spending most of my time thinking and using AI to express those thoughts? Then I'm a developer who uses AI.

Am I spending most of my time prompting and occasionally skimming? Then I'm a prompt engineer who used to be a developer.

The terrifying part is that the ratio shifts quietly. You don't notice it moving until someone asks a simple question and you don't have an answer.


What I'm Actually Doing About It

I'm not quitting AI. That's not the answer, and honestly it's not what I want. AI has made me faster at the parts of development I find least interesting, which in theory should free me up for the parts I find most interesting.

The problem is that "in theory" is doing a lot of work in that sentence.

So I'm trying small things. Not a productivity system. Not a manifesto. Small things.

One hour, no AI, every morning. The first hour of my coding day — no Copilot, no Cursor, nothing. Just me and the problem. It's slower. Sometimes frustrating. It's also mine in a way that the rest of the day often isn't.

One honest question at the end of each day. "Did I think today, or did I just generate?" No audience. No performance. Just an honest answer to myself.

Building things nobody will ever see. No metrics. No deployment. No PR approvals. Just creation for the sake of creating, which turns out to be harder than it sounds when you've spent years optimizing for output.

Remembering the junior's question. Not to feel guilty. To stay honest about the answer.

Will these things fix the identity crisis? Probably not. But they slow the drift. And right now, slowing the drift feels like enough.


The Hard Truth

Here's what I've accepted: I'll never be the developer I was before AI. That version of me is gone not because AI took something from me, but because I gave it away. One shortcut at a time. One skipped debugging session at a time. One prompt where there used to be thinking.

But I don't think that makes me just a prompt engineer.

It means I need a new, honest answer to the question. One that accounts for what I've lost and what I've actually gained. One that doesn't pretend the craft is exactly what it used to be, but doesn't write it off either.

Developer who uses AI feels close.

Developer who still cares about the difference feels closer.


One Question Before You Go

What do you call yourself now? Developer, prompt engineer, something in between, something you're still figuring out?

And more importantly does the title actually matter, or is it only the work that does?

I've been thinking about this for weeks and I still don't have a clean answer. I'd genuinely like to hear yours.

I'll go first in the comments.

Your turn. 👇


The junior developer conversation is real. I used AI to help structure my thoughts for this which is either ironic or exactly the point.