In This Guide:
The Right Mindset for AI Writing
Most people approach AI writing wrong. They expect to type one vague prompt and get a finished, publishable piece. That's not how it works — and that's not even what you want.
The right mental model: AI is your co-writer. You bring the ideas, the context, the judgment, and the final editorial eye. AI handles the drafting, the structure, the options, and the speed. Together, you produce better work faster than either could alone.
The practical reality: The best AI writing workflow is about 30% AI effort and 70% human editorial judgment. AI saves you the hardest part — staring at a blank page — but your voice, your knowledge, and your edits are what make the output actually good.
Emails — The Best Place to Start
Email is where most people first experience the "oh wow" moment with AI writing. It's perfect for AI because: there's a clear purpose, there's an audience you can describe, there are constraints (length, tone), and you know immediately whether the output is right.
Drafting from scratch
Write a [tone] email to [recipient/role] about [topic]. The key points are: [1-3 bullet points]. Length: [short/medium/under X words]. Goal: [what should happen after they read it].
Write a professional but warm email to a potential client named James Chen at Apex Marketing. I'm following up on a meeting last week where we discussed social media management. Key points: we can start immediately, our rate is $2,500/month, I'd like to schedule a 30-min next-steps call. Keep it under 150 words.
Handling difficult emails
AI is exceptionally good at defusing difficult situations — complaints, rejections, delays — because it drafts responses that are firm but not confrontational, and empathetic without being a pushover.
Help me write an email declining a job offer graciously. I want to leave a good impression and keep the door open for the future. The job was [role] at [company], and the main reason I'm declining is [brief reason — salary/other offer/fit].
Blog Posts & Long-Form Content
AI is excellent at generating blog post structures and first drafts. The key is giving it enough context about your audience and angle — then doing real editorial work on the output.
The 3-step blog workflow
- Get an outline first — Ask AI to outline the post before writing it. Review and adjust the outline to match what you actually want.
- Write section by section — Paste the outline and have AI write one section at a time. This gives you more control than asking for the whole post at once.
- Add your examples and voice — Replace generic examples with your real ones. Add your opinions. Remove anything that sounds templated.
Create an outline for a 1,000-word blog post titled "[your title]". Target reader: [describe audience]. The post should: [goal — inform/persuade/entertain]. Include: intro hook, 3-5 main sections with subpoints, and a conclusion with a call-to-action.
Social Media Content
Social media is one of the highest-value AI writing use cases because: posts are short (easy for AI), volume matters (you need many of them), and variation matters (same content, multiple formats).
Write 5 LinkedIn posts about [topic/experience]. Each post should: be 150-250 words, start with a hook that creates curiosity, share one specific insight or lesson, end with a question to drive comments. Make them feel personal and real, not corporate.
Power move: Ask AI to repurpose one piece of content into multiple formats: "Take this blog post and create: a LinkedIn post, a Twitter thread (5 tweets), and a short email newsletter section."
Using AI to Edit Your Writing
This might be more valuable than using AI to write from scratch. Paste your existing draft and ask for specific improvements:
- "Make this opening paragraph stronger — it needs to create more curiosity"
- "Simplify this — my audience is non-technical"
- "This is too long. Cut it to 200 words without losing the key message"
- "The tone feels too formal. Make it conversational without being unprofessional"
- "What's the weakest part of this argument? How would you address it?"
How to Make AI Sound Like You
The biggest complaint about AI writing: "It doesn't sound like me." Here's how to fix that:
- Show it your writing first — Paste 2-3 examples of your past writing and say "This is my writing style. Match this voice when you write for me."
- Be specific about tone — "direct and slightly sarcastic", "warm and encouraging", "professional but approachable" — specific is better than generic
- List words/phrases to avoid — "Don't use the word 'delve'. Don't use 'crucial'. Don't start sentences with 'In today's world'"
- Add your examples — Replace AI's generic examples with real ones from your experience
- Edit the opening and closing — These are the most "AI-sounding" parts. Rewrite them in your own words.
5 Common Mistakes to Avoid
❌ Vague prompts
"Write a blog post about marketing" produces generic output because there's no context about your angle, audience, or goal.
✓ Specific prompts
Include audience, goal, tone, length, and key points. The more specific your prompt, the less editing you need to do.
❌ Publishing without editing
AI output needs a human editorial pass. Publish it raw and your readers will know it was AI — and won't engage.
✓ Edit for your voice
Treat AI output as a draft. Swap in your examples, adjust the tone, and add your genuine opinion.
❌ Accepting facts without verifying
AI will include statistics, quotes, and specific claims that may be incorrect. Never publish factual claims without verifying them.
✓ Fact-check before publishing
Verify any statistics, quotes, or specific claims. Ask AI to note where it's uncertain.
The 10-minute edit rule: Budget at least 10 minutes of editing for every 500 words of AI output. Not because AI writing is bad — but because editing is what makes it yours.