It was 11:47 PM on a Tuesday, and I was still awake replying to client emails, formatting an invoice for the third time because I kept fat-fingering the totals, and trying to remember which client wanted the blog post in British English versus American English.
I hadn't written a single word of actual paid work in four hours.
This was my life for the better part of two years as a freelancer. Not the highlight reel you see on Instagram—the "quit my 9-5, sip coffee on my laptop by the beach" version. The real version: administrative chaos, undercharging because I was afraid to negotiate, and burning out on tasks that had nothing to do with my actual skills.
If you're a freelancer reading this, I'm guessing at least one of those things sounds familiar.
My Situation: Freelancer by Day, Aspiring Coder by Night
Here's where I was at: I'd been freelancing for a few years—content writing, some light design work, a bit of consulting. I was making decent money, but I was also drowning. Every hour I spent on admin was an hour I wasn't spending on billable work or, frankly, sleeping.
On top of that, I'd started teaching myself to code. Not because I wanted to become a full-time developer, but because I kept hitting walls in my freelance work that a little technical knowledge could solve—automating repetitive tasks, building simple websites for clients instead of outsourcing it, understanding enough to troubleshoot when things broke.
The problem? Learning to code also takes time. And I didn't have much of that to spare.
So there I was: overworked freelancer, frustrated beginner coder, googling "how to stop feeling like I'm failing at everything" at midnight.
The Turning Point
I wish I could tell you I had some genius epiphany. I didn't. A friend (also a freelancer) mentioned offhand that she'd started using ChatGPT to draft client proposals and it cut her writing time in half.
I was skeptical. I thought AI tools were either gimmicky or going to produce robotic, generic garbage that I'd have to rewrite anyway.
I tried it anyway, because midnight invoice formatting had officially broken me.
The first thing I did was paste in a messy, rambling brain-dump of ideas for a client pitch and ask ChatGPT to organize it into a clear proposal structure. It took me 90 seconds. What used to take me 45 minutes of staring at a blank Google Doc took less than two.
That was the moment something clicked. This wasn't about replacing my skills or my voice—it was about removing the friction between "I know what I want to say" and "I've actually said it clearly."
From there, I started experimenting. A lot. And slowly, my freelance business (and my coding journey) started looking completely different.
How AI Saved Me Time (Real Numbers, Not Hype)
Let's start with the obvious one, because time is the thing freelancers never have enough of.
My actual workflow now: Every Monday, I use ChatGPT to help me draft my weekly client check-in emails. I give it bullet points of what I did for each client that week, and it turns them into polished, professional updates. What used to take me an hour across five clients now takes about 15 minutes.
I also started using Notion AI to summarize long client calls I'd recorded and transcribed. Instead of re-listening to a 45-minute call to extract action items, I paste the transcript in and ask for a summary with clear next steps. That alone saves me 30-40 minutes per client call.
Here's a prompt I use constantly for email drafts:
"Turn these bullet points into a professional but friendly client update email. Keep it concise, highlight progress, and end with a clear next step or question."
Total time saved per week? Conservatively, 5-7 hours. That's almost a full workday I've gotten back—not by working faster, but by cutting out the friction of starting from a blank page.
How AI Helped Me Earn More
This one surprised me. I didn't expect AI to directly impact my income, but it did—in two specific ways.
First: better proposals, faster. I used to dread writing proposals because they took forever and I never felt confident about my pricing language. Now, I use AI to help me draft the first version of a proposal, then I edit it to sound like me. This means I can pitch more potential clients in the same amount of time, which means more opportunities in the pipeline.
Second: I started offering new services I couldn't have offered before. Because I was learning basic coding with AI's help (more on that below), I started offering simple website tweaks and automation setups to existing clients—things like connecting their contact forms to email lists or cleaning up basic HTML/CSS issues on their sites. These weren't huge jobs, but they added an extra $300-600 a month from clients I already had relationships with. I didn't have to find new clients; I just expanded what I could offer to the ones I already had.
Here's the prompt I use when drafting a new service offering:
"I want to offer [specific service] to my existing clients. Help me write a short, non-salesy email introducing this new offering, explaining the benefit in plain language."
How AI Improved My Work Quality
I was worried, at first, that leaning on AI would make my work feel generic. Turns out, the opposite happened—because I stopped spending mental energy on the boring stuff and had more left over for the parts that actually require creativity and judgment.
For content writing, I use AI as a first-draft generator or an editor, never as the final voice. I'll write my own draft, then ask AI: "Review this for clarity and flow, and flag any sentences that feel repetitive or confusing." It catches things I miss because I'm too close to my own writing.
For client-facing documents (contracts, style guides, onboarding docs), I use AI to check for gaps. A prompt I use often:
"Review this freelance contract and tell me if there are any standard clauses I'm missing that protect me as the freelancer."
It's not a replacement for a lawyer, but it's caught things I'd overlooked—like unclear payment terms—that I then fixed before sending things out. That's real quality improvement, not just speed.
How AI Is Helping Me Learn to Code, Faster Than I Expected
This might be the part that surprised me most.
Learning to code alone is intimidating. You hit an error message, you have no idea what it means, and you spend 40 minutes on Stack Overflow feeling stupid. I've been there more times than I can count.
Now, when I hit an error, I paste it directly into ChatGPT or GitHub Copilot (which I use inside my code editor) and ask what's going wrong. Instead of feeling lost for 40 minutes, I get an explanation in plain English within seconds—often with a suggested fix I can test immediately.
I also use AI as a patient, judgment-free tutor. When I don't understand a coding concept, I ask AI to explain it "like I'm a complete beginner" or "using a real-world analogy." That reframing has helped concepts click that textbooks never managed to explain clearly.
A prompt I use almost daily while coding:
"Explain what this code does line by line, as if I'm a beginner. Then tell me what would happen if I changed [specific part]."
I'm not going to pretend I'm suddenly a expert developer. I'm not. But I've gone from being completely stuck and discouraged to actually building small, functional things for my freelance clients—which has been both a confidence boost and a genuine revenue add.
Practical Takeaway: What You Can Do TODAY
You don't need to overhaul your entire workflow overnight. Here's what I'd actually recommend starting with:
Pick one repetitive task you hate (client emails, proposal writing, invoice reminders) and try using ChatGPT or Claude to draft it for you this week. Just one task. See how much time it saves.
Use AI as an editor, not a writer. Write your own first draft of something, then ask AI to review it for clarity, tone, or gaps. This keeps your voice intact while catching blind spots.
If you're curious about coding, start small with AI as your tutor. Don't try to learn everything at once. Pick one tiny project (a simple webpage, a basic automation) and use AI to explain each step as you go.
That's it. Small, specific, doable today—not a 10-step system that requires you to reorganize your entire business.
The Reality Check
I want to be honest with you, because I think a lot of AI content online oversells this stuff.
AI is not magic. It doesn't understand your clients, your niche expertise, or the nuance of your specific industry the way you do. I've gotten plenty of generic, slightly-off responses that needed heavy editing. Sometimes it's genuinely faster to just write something myself.
It also doesn't replace actual expertise. AI helped me learn to code faster, but it didn't learn for me—I still had to put in the hours, make mistakes, and build understanding. Same with freelance work: AI drafts a proposal, but I still need to know what makes a good pitch in the first place.
And honestly? There's a learning curve to using these tools well. My first few weeks of prompts were clunky and the outputs reflected that. It got better as I got more specific and understood each tool's strengths.
This isn't about replacing your skills or your judgment. It's about removing friction so you can spend more time on the parts of your work that actually require you.
Let's Figure This Out Together
I'm not an AI expert. I'm not a coding expert either. I'm a freelancer who got tired of drowning in admin work and decided to experiment—messily, imperfectly, one prompt at a time.
If you're curious about using AI in your freelance work but don't know where to start, or if you're on your own coding journey and want to see how someone else is figuring it out in real time, come along for the ride.
I write about this stuff every week on my Substack—real experiments, real prompts, real numbers (including the failures, because there are plenty). No hype, no "10x your business overnight" nonsense. Just honest updates from someone building this alongside you.
Subscribe here, and let's figure out how to work smarter—not harder—together.
Top comments (0)