How AI-assisted blogging became a $7K/month side hustle

Everyone assumed AI would wipe out writing jobs. Instead, demand for good writers is climbing again – especially for those who know how to work with AI, not compete with it. Today, tech companies are happily paying $900–$1,200 per long-form article to writers who can blend real expertise, original research, and AI assistance.

In this guide, you’ll see exactly how a non-technical writer built a $6K–$7K/month freelance business writing AI-assisted blog posts for SaaS companies – and how you can follow a similar path even if you don’t have a writing degree or coding background.

Why AI didn’t kill writing – it raised the bar

Search data tells a surprising story. Interest in the phrase “hire a writer” dipped when ChatGPT became mainstream in 2023, but then rebounded and climbed higher than ever. Companies tried to replace writers with AI, discovered that generic AI content doesn’t perform, and started coming back to human writers who can use AI well.

What changed is not whether writers are needed, but what kind of writers are in demand. Businesses now want people who:

• Understand their tools or industry well enough to write with authority
• Use AI to move faster and think better, not to churn out bland content
• Bring personal stories, cultural awareness, and original angles that AI can’t fake

That’s where AI-assisted writers are thriving.

Meet the $1,000-per-article AI-assisted writer

Stephanie is a freelance writer based in Austin who focuses on long-form content for tech and B2B SaaS companies. She charges $900–$1,200 per blog post and typically writes 2,500–3,000-word articles that are designed for SEO but still enjoyable for real humans.

She’s not a developer. She has no coding background. Her work is a mix of:

• Long-form blog posts and tutorials for SaaS tools like Zapier, Jotform, and Softr
• Technical-style guides and feature explainers written for non-technical readers
• Side projects like coaching, editing résumés, and helping students with college essays

All in, she currently brings in around $6,000–$7,000 per month, mostly from writing. The twist: she uses AI heavily – but not the way most people think.

The AI stack she actually uses

Instead of relying on a single model, Stephanie treats AI tools like a toolbox, using each for what it does best:

• ChatGPT – for general writing help and research
• Claude – for marketing content and long-form drafts that match her voice more closely
• Gemini – for processing video and audio (transcription, summarizing, extracting notes)
• Cursor – as the central workspace that routes tasks to different models (e.g., send video to Gemini, send notes and templates to Claude)

She also uses specialized tools like Granola to capture and summarize client calls, and she’s deeply familiar with automation platforms like Zapier from her full-time role there.

How she builds AI-assisted tutorials and guides

One of Stephanie’s most effective workflows is for writing detailed, step-by-step tutorials about automation workflows. Here’s how she does it:

1. Capture the real thinking process

Instead of sitting down to write from scratch, she records her screen (using a tool like Loom) while she actually builds the workflow she’s going to write about. She narrates everything she’s doing and thinking:

• Where she hesitates or gets confused
• Why she clicks a certain option
• What feels tricky or non-obvious

She calls this “writing is thinking.” The goal is to capture her first reactions and real problem-solving, because those are the gold nuggets that make an article genuinely helpful.

2. Use Gemini to process the video and audio

Next, she uses Cursor to send that recording to Gemini with clear instructions in plain English, such as:

• Transcribe the audio and summarize her narration
• Detect points where she hesitates, hovers, or seems unsure
• Extract those “sticky” steps and turn them into detailed notes

Gemini is especially good at handling video and audio, so it becomes her “thinking extractor.”

3. Use Claude to turn notes into a draft

She then takes the structured notes and a pre-written article template (for example, a standardized tutorial layout) and feeds them into Claude via Cursor. Claude uses her template plus the notes to produce a strong first draft.

The result is not a finished article, but a well-structured, accurate draft that already includes the tricky details she surfaced while building the workflow.

4. Add her voice and fix the details

Finally, she edits the draft herself. She:

• Corrects anything AI misunderstood
• Adds personality, metaphors, and stories
• Punches up bland sections and clarifies confusing steps
• Checks screenshots, UI labels, and product changes

That human editing pass is where the article stops sounding like a generic AI tutorial and starts sounding like something only she could write.

How much AI is in a $1,000 article?

For a typical 3,000-word blog post, Stephanie estimates that:

• About 70–80% of the words are written or heavily rewritten by her
• The underlying structure, angle, and ideas are almost always 100% hers
• AI mainly helps with summarizing, outlining, and turning raw material into a usable draft

Even when she uploads dozens of her past articles to a model like Claude to mimic her style, it still can’t fully replace her. It doesn’t have her memories, lived experiences, or spontaneous associations – like comparing Zapier MCP to an action role-playing game, or opening a tutorial with a childhood story that illustrates a point.

AI can imitate tone, but it can’t invent your life. And if you let AI do all your writing, your own voice stops evolving. You stay static while the market keeps rewarding writers who are actively sharpening their skills.

Why tech companies pay $900–$1,200 per post

From the outside, $1,000 for a blog post sounds high. From the inside of a SaaS company, it’s often a bargain.

Here’s how they think about it:

• SEO and discovery: A well-written comparison or “alternatives” article can rank for high-intent keywords and bring in buyers for years.
• Authority and trust: Tutorials, feature guides, and explainers help position the product as the expert solution in its category.
• Speed: When something changes in the ecosystem (a competitor shuts down, a new feature launches), they need content that day – and AI-assisted writers can move faster without sacrificing quality.
• Internal bandwidth: Marketing teams are often overloaded. Outsourcing long-form content to a proven freelancer is cheaper than hiring another full-time employee.

Companies also know that pure AI content is risky. Platforms like YouTube are demonetizing faceless AI-generated channels, Snapchat is pushing back on AI videos, and Anthropic is watermarking AI-generated text. Businesses are realizing they can’t just flood the web with AI sludge and expect good results.

They want writers who understand tools like Claude, Gemini, and ChatGPT – but who also bring real expertise and originality. That’s the sweet spot.

What she actually does during those 5–6 hours

For each 2,500–3,000-word article, Stephanie spends about five to six hours. That time is not just “typing words into a document.” It typically includes:

• Hands-on tool testing: Signing up for the SaaS products she’s writing about, clicking around, trying features, and understanding how they really work.
• Research: Reading Reddit threads, community posts, and user reviews to see what real users love or hate (pricing confusion, UX issues, hidden features).
• Outline and angle: Deciding what the article should emphasize – for example, positioning one tool as a better alternative for a specific use case.
• AI-assisted drafting: Using her templates plus AI to get a strong first draft quickly.
• Editing and polishing: Injecting her voice, adding examples, clarifying steps, and making sure the piece is genuinely useful and on-brand.

She also follows SEO best practices: solid headings, clear structure, no keyword stuffing, and content that actually answers what searchers are trying to do. Increasingly, she also thinks about “GEO” – generative engine optimization – making content that’s structured and clear enough to be surfaced by AI assistants as an answer.

How she built her freelance client base

Stephanie didn’t jump straight into $1,000 articles. Her path started in 2016, and it’s very repeatable today:

• She joined Upwork early and got her first clients there.
• She took on a mix of projects: blog posts, website copy, editing, and even PR work.
• She worked with interesting clients, including the son of Biggie Smalls (Notorious B.I.G.), helping with a CBD brand’s communications.
• She later landed contracts with Apple (editing sales enablement materials) and Zapier (full-time writing and editing).

Even now, with years of experience, she still uses Upwork and considers it underrated for beginners. Clients are already there looking for writers, which removes the hardest part when you’re starting: finding leads.

Why you don’t need a writing degree to do this

To prepare for teaching others, Stephanie interviewed several successful freelance writers on Upwork who earn at least $3,000 per month. None of them had a formal writing degree.

What they all did have:

• A small but focused portfolio (4–5 strong samples)
• A clear niche or type of writing they specialized in (e.g., SaaS blogs, e-commerce copy, or technical explainers)
• A basic understanding of SEO and online content formats
• The willingness to pitch, get rejected, and keep going

In other words, you don’t need credentials. You need proof.

Step 1: Pick your niche and understand the landscape

If you want to start in the next 90 days, don’t try to be a “writer for everything.” Narrow down.

Look at the main buckets of work out there:

• Creative and publishing: Ghostwriting books, editing fiction or memoirs, helping creators write newsletters.
• Marketing and business: Blog posts, landing pages, email sequences, ad copy, scripts for videos or webinars.
• Tech and SaaS: Tutorials, feature guides, comparison posts, API documentation, automation workflows.

Then ask: what do you actually like? If you’re obsessed with barbecue, you could write for grill brands, restaurants, or e-commerce stores selling BBQ tools. If you’re into AI tools, you can focus on SaaS and automation content and build a career similar to Stephanie’s. For more inspiration on AI-focused side hustles, you might like this breakdown of real ways to make money with AI tools.

Step 2: Build a simple but strong portfolio

Before clients will pay you, they need to see what you can do. You don’t need big brand logos – you just need good samples.

Here’s a simple way to build that in a few weeks:

• Write 3 blog posts in your niche (e.g., SaaS tutorials, BBQ guides, AI tool comparisons).
• Add 1 landing page or sales page to show you can write conversion-focused copy.
• Optionally add 1 more format (like an email sequence or a short case study) to show range.

You don’t have to wait to get hired. If you love a brand, write a sample article as if they were your client. For example, “What makes Franklin Barbecue different” could be a great sample piece. You can even send it to them later as a warm introduction.

Host your portfolio on a simple website if you can – it looks more professional than a bare Google Drive folder. But if a website is a blocker, a clean shared folder is better than nothing.

Step 3: Learn to pitch like a human

Clients are everywhere and nowhere – they’re on LinkedIn, Upwork, in niche communities, and behind the brands you already follow. The skill that unlocks them is the cold pitch.

A good pitch is short and specific. In 4–5 sentences, you should:

1. Briefly introduce who you are and what you do.
2. Show that you’ve actually paid attention to their brand (reference a podcast episode, a specific product, or a recent post).
3. Explain how you can help in a concrete way (e.g., “I specialize in long-form comparison posts for SaaS tools and can help you rank for X, Y, Z keywords”).
4. Link your portfolio.
5. Invite a quick call or ask if they’re open to freelance help.

Stephanie once listened to an entire podcast from a marketing agency before pitching them. She referenced a specific detail in her message, and they were so impressed they called her immediately. A little extra effort can stand out in a crowded inbox.

Step 4: Use AI to speed up, not to replace yourself

One of the biggest mistakes new writers make is trying to automate everything with AI. They spend more time configuring tools than actually writing, and the output still needs heavy fixing.

Instead, use AI in the 1–2 parts of your workflow where it helps most:

• Summarizing research and calls
• Turning raw notes or transcripts into an outline
• Drafting sections that you will then heavily edit
• Checking for gaps, inconsistencies, or missed edge cases

This “human in the loop” approach mirrors what successful companies are doing internally. In one Zapier report Stephanie edited, they found that teams who used AI effectively only used it for about 18% of each workflow – usually one or two steps, not the entire process.

If you want a deeper dive into a specific AI-powered side hustle, check out this guide on a Claude-based side hustle that also leans on the human-in-the-loop model.

Step 5: Price your work with confidence

When you’re starting out, pricing feels like guesswork. Stephanie suggests a few practical guidelines:

• On platforms like Upwork, many solid beginners are charging around $30–$40 per hour.
• Clients may negotiate down; if the project offers good experience or portfolio value, a slightly lower rate can still be worth it early on.
• For fixed-price projects (like a full website rewrite), estimate your hours based on scope, then quote a project fee instead of hourly. This avoids surprises for both sides.

To scope a project, she usually asks:

• How many pages or words are we talking about?
• Is there existing copy, and how good is it?
• How much input will I need from the client (e.g., calls, internal docs)?

On calls, she often uses a tool like Granola to record and transcribe the conversation, then queries it later for details – another simple way AI saves time without touching the actual writing.

What her income looks like now

After leaving her full-time role at Zapier, Stephanie’s freelance income currently looks like this:

• 5–10 blog posts per month for tech/SaaS companies at $800–$1,200 each (most around $900–$1,200).
• Occasional student packages for college application essays, sold in bundles of five sessions at $90/hour ($450 per package).
• Coaching and editing for executives and other writers, with variable package pricing.

All told, she’s at roughly $6,000–$7,000 per month – and she only left her full-time job a few weeks ago. Most of that income is from writing technical-leaning content as a non-technical person, powered by smart use of AI.

The real opportunity: being the human AI can’t replace

We’re already seeing a pattern: companies who tried to replace writers with AI are quietly coming back to freelancers a year later. Platforms are cracking down on low-effort AI content. AI providers like Anthropic are watermarking generated text. Even Anthropic’s own CEO has said he uses Claude for research and assistance, not for full, finished writing.

In that environment, the writers who win are those who:

• Embrace AI as a collaborator, not a crutch
• Do real research and hands-on testing
• Bring stories, analogies, and cultural awareness AI doesn’t have
• Keep practicing the hard part of writing – thinking clearly on the page

Learning to stay focused, be articulate, and turn complex tools into clear, engaging content is exactly why clients are paying $900–$1,200 per article. If you’re willing to build a portfolio, pitch bravely, and use AI wisely, there’s plenty of room for another AI-assisted writer earning $7K a month.