The First Time I Heard About Prompt Engineering
I had no clue what a “prompt engineer” even was until mid-2023.
I saw a LinkedIn job post—something about “writing stuff for AI” with a salary way higher than anything I expected for, well, writing stuff. I laughed. Closed the tab. Thought it was just another tech buzzword that would fizzle out in a month.
Fast forward to 2025: it’s not only a real job—it’s my job.
And honestly? It’s the most fun I’ve had in tech since I built my first blog in college.
The path here wasn’t clean. It was confusing, chaotic, unstructured—and somehow, that made it feel right.
If you’re curious about how to get into prompt engineering (like actually doing it—not just reading about it on Twitter), here’s my honest take: how I broke in, what skills mattered most, what I actually do, and why this work is way more than just asking ChatGPT to write poems.
It Starts with Curiosity—Not Code
Let’s bust a myth: you don’t need to be a full-stack wizard to become a prompt engineer.
Sure, I know just enough Python to break things gently, but prompt engineering is more about how you think than what you code.
At first, I thought writing prompts was just about saying “please” and being polite to an AI. But it’s not that.
It’s instruction design. Communication theory. Behavioral nudging. Reverse-engineering how an alien brain processes language.
You feed it words. It feeds you results. If the results suck? That’s on you.
It’s part logic, part psychology, part storytelling—and weirdly satisfying.
What Helped Me Get (Kinda) Good at This
Here’s what made the biggest difference for me early on:
1. Writing. Just… a lot of writing.
Tweets. Blog posts. Rants. Practicing how to express ideas clearly helped me become 10x better at shaping AI behavior.
2. Treating AI like a creative partner, not a tool.
I started chatting with ChatGPT like it was my co-worker. “Here’s what I want. Here’s what I don’t want.” The more context I gave, the better the results got.
3. Trying 50 variations of the same prompt.
I once tested 47 different versions of a single product description prompt. Brain-melting. But I learned patterns.
4. Reverse-engineering great outputs.
When something came out perfectly, I’d break it down. What made the prompt tick? Could I reuse the structure? Modify the tone?
Tools I Use Daily as a Prompt Engineer
You might be surprised how simple the toolkit is:
- ChatGPT, Claude, Gemini, Mistral — for testing and comparing behavior
- Excalidraw / Whimsical — for visualizing multi-step prompt chains
- Notion / Obsidian — where I log all my prompt experiments (yes, I keep a prompt journal)
- Coffee & noise-canceling headphones — crucial for deep “AI conversations”
That’s it. No complex IDEs. Just fast feedback loops, lots of logs, and creative problem-solving.
What My Job Actually Looks Like
This isn’t just about writing prompts like “Write me a funny poem about broccoli.”
Some days, I’m crafting system prompts for customer service bots—like designing how the AI should handle angry users.
Other days, I’m building prompt chains that help AI tutors feel like real teachers, or improving internal knowledge tools with smart summarization workflows.
But here’s the kicker: most of my job is editing.
Not writing from scratch. Just tweaking, refining, stress-testing, making prompts more inclusive, more robust, and less likely to produce garbage.
It’s kind of like UX design. You’re designing a conversation, not an interface.
What They Don’t Tell You
- Prompt engineering is exhausting.
You can spend hours on a prompt that almost works… then the model updates overnight and breaks everything. - It’s hard to explain to non-tech folks.
“So… you talk to robots for a living?” Yeah, kinda. - Imposter syndrome is real.
Are you good at this? Or are you just a really persistent guesser?
Can You Actually Get Hired as a Prompt Engineer?
Yes.
But you’ll need to show that you understand models—how to shape behavior, optimize language, and experiment thoughtfully.
Some companies post prompt engineering roles under different names:
- AI Interaction Designer
- AI Content Strategist
- LLM Behavior Architect
- Conversational UX Specialist
Read the job descriptions closely.
I got my first gig with a portfolio.
Literally a Notion page full of prompt examples, outputs, screenshots, and notes. Nothing flashy. But it showed how I think—and that’s what got me in.
Read more about tech blogs . To know more about and to work with industry experts visit internboot.com .
Final Thoughts: Just Start
If you’re curious about prompt engineering in 2025, don’t wait for a course. There’s no playbook. You’re helping write it.
- Talk to AI every day.
- Experiment with tone and structure.
- Break it. Fix it. Repeat.
- Log everything.
- Get weird.
- Stay curious.
This job is weird. It’s creative. Sometimes lonely. But it’s also amazing.
Because you’re not just asking questions—you’re shaping how machines think.
And that’s wild.
