Create a reusable voice profile by feeding AI high-signal writing samples (customer emails, internal messages, transcripts, prior posts) plus explicit style preferences (sentence length, words to avoid, signature opinions). Use it as persistent context so drafts match the founder’s authentic voice instead of generic marketing copy.
The human provides the real inputs (opinions, experience, examples) and defines what “sounds like us.” They decide what is true, what is publishable, and which viewpoints to stand behind.
Upload 15–20 past posts/emails and have AI distill them into a concise voice profile (tone, structure, banned phrases, signature takes) that you reuse for every draft.
ChatGPT (Custom Instructions/GPTs) / Claude (Projects)Create a brand voice using existing content samples so the tool can generate LinkedIn posts that consistently match your preferred cadence and phrasing.
Jasper (Brand Voice) / Copy.ai (Brand Voice)Store do/don’t rules and best-performing post examples in a shared doc/knowledge base that AI can reference whenever generating new content.
Notion AI / Google Docs + GeminiGenerate a draft in your style, edit, and schedule it in the same workflow to maintain consistent posting cadence.
Taplio✅ How I use AI for marketing:
1. Build dedicated AI projects per marketing function
2. Train each project on your actual data and brand voice
3. Use AI for research and analysis before creation
4. Generate content at scale with programmatic templates
5. Come up with creative, scalable digital PR ideas
6. Fact-check my own bullshit when I'm not at 100% capacity
...
I use AI for about 70% of my LinkedIn content.
(and people still say it sounds like me)
Here's the thing most people get wrong:
They prompt ChatGPT with a topic.
Get a draft back.
Hit post.
Then wonder why it sounds like everyone else.
AI can write the whole post.
That's not the problem.
The problem is does the AI know you?
Here's the 3-part filter that makes AI content sound human:
𝟭) 𝗣𝗿𝗼𝗼𝗳 𝗼𝗳 𝘄𝗼𝗿𝗸
↳ Real numbers. Real screenshots. Real outcomes.
↳ If you haven't lived it, don't post it.
𝟮) 𝗘𝗺𝗼𝘁𝗶𝗼𝗻
↳ What did you feel when it happened?
↳ Frustration? Relief? Surprise?
↳ That's what makes it human.
𝟯) 𝗣𝗢𝗩
↳ What's your take that others might disagree with?
↳ A clear opinion beats a safe summary every time.
AI can give you structure.
AI can give you speed.
Stanley is my secret weapon. He gets to know me and understands my lived experience. He understands how I feel about things, and he can weave in the human side to my posts.
P.S. What's your biggest struggle with AI content - sounding human or finding ideas?
I’ve been running a B2B SaaS company and content marketing is my biggest inbound channel—but it used to take a lot of time. I tried generic ChatGPT prompts, Jasper/Copy.ai templates, and even a freelancer + AI, but everything sounded bland and “AI-written.” What finally worked was building a “voice profile” from 15–20 real writing samples (emails, Slack messages, podcast transcripts, past LinkedIn posts) and using that as context so AI could draft content in my actual voice. My workflow now is: record short voice notes → AI drafts 3–5 posts per note → I quickly review/edit/reject → schedule posts and respond to comments. This increased my posting volume from 2–3/month to 15–20/month while cutting time spent from ~8–10 hours/month to ~2–3 hours/month, and improved average impressions and inbound DMs.