Why your AI content sucks, and how to bring back the human touch

Lately, I’ve noticed a change in content from companies I’ve been following for years. Usually, their content is insightful, witty, and feels like they were written just for me. But lately? It reads like it’s been churned out by ChatGPT after being prompted with “write a post about productivity tips”.
Generic. Bland. Soulless.
I unfollow immediately. Cause if I want generic tips, I might as well prompt ChatGPT myself.
Here’s the thing: I know AI was involved in creating that content. And I’m not against using AI tools—heck, I use them daily in my own content creation process. But there’s a massive difference between using AI as a tool and letting it take over your entire content creation.
If you’re reading this, chances are you’ve already integrated AI into your content workflow. Maybe you’re even wondering if your content has that same artificial aftertaste I’ve been seeing. Let me tell you: if you’re worried about it, your audience probably notices it too.
The problem with pure AI content
“It’s going to seem good because it seems coherent. But coherent doesn’t necessarily mean good, right?”
The telltale signs of AI-only content
You know it when you see it. Content that’s technically correct but oddly generic. Paragraphs that seem to say a lot without actually saying anything meaningful. An absence of original insights or perspectives.
The worst offenders typically have:
- Unnecessary formality (”Individuals may find that...”)
- Overly balanced perspectives that never take a real stance
- Generic examples that could apply to literally any business
- A conspicuous lack of industry-specific insights or terminology
- The infamous “In conclusion...” paragraph that basically repeats the introduction
I call this empty calorie content—it fills the page but provides zero nutritional value to your readers.
Why readers can spot it (and hate it)
Here's what most marketers miss: your audience is getting increasingly sophisticated at spotting AI-generated content. They might not explicitly think, “This was written by a bot”, but they definitely feel something is off.
People don’t just read content—they feel it. When there’s no human energy behind the words, no personality infused into the paragraphs, readers sense the emptiness. And that creates an immediate disconnect with your brand.
Most marketers learn this the hard way after publishing fully AI-generated articles. The metrics look fine—decent time on page, okay scroll depth—but the comments and social shares? Crickets. No engagement. No one feels compelled to respond because the content doesn’t feel like it came from a real person.
How it damages your brand credibility
When everything you publish sounds like it could have been written by anyone (or anything), you’re not building brand authority—you’re diluting it.
Think about it: your best content should showcase your unique expertise, perspective, and voice. It should demonstrate why someone should listen to you instead of the thousands of other voices in your space.
AI-only content does the exact opposite. It homogenises your brand voice and makes you sound like everybody else.
How AI content affects SEO and search performance
Search has changed dramatically with generative AI, and it’s impacting how our content gets found. You can’t just “check the SEO box” anymore and expect results.
With GenAI search engines, we’re dealing with a completely different beast. As I’ve discovered through working with clients, these engines prioritise context and authority above all else.
Here’s the irony: the generic AI-generated draft that seems “good enough” is actually too generic to appear in AI search responses. To stand out, you need to make sure that whatever draft you get from AI is only about 80% of your final content. That remaining 20%? That's where you need to add authority-building elements that AI can’t create:
- Expert insights (both internal and external)
- Original quotes that demonstrate real expertise
- Multimedia elements and visuals that help communicate your message
- Data and research that backs up your claims
I haven’t had to change my strategy much with clients because of generative search results when they were already structured with authority building and actionable insights. Their content started to appear in results because they were already following best practices. But the bare minimum “good enough” content? It's completely invisible in today's search results.
Adapting content strategy for AI search vs. traditional SEO
The strategies that worked for traditional SEO simply don’t translate directly to AI search. Let me give you a concrete example.
In traditional SEO, you could answer a question like "What is a startup?" with two simple sentences and potentially appear in a featured snippet. That strategy worked fine for years. But with generative AI search, that approach falls completely flat.
The difference is fundamental: generative AI search prioritises context and authority in a way that traditional search didn’t do as much. It doesn’t just want the answer—it wants comprehensive understanding.
What this means for your content strategy:
- Depth over breadth: Instead of creating 20 thin articles, create 5 comprehensive guides.
- Authority signals matter more: Include quotes from recognised experts, cite reputable sources, and demonstrate genuine expertise.
- Contextual relevance: Your content needs to address not just the primary question but the related questions and concerns someone might have.
- Actionable value: AI search gives preference to content that doesn't just inform but enables action.
This shift requires us to raise our content quality bar significantly. The good news? If you’re already creating genuinely valuable content, you’re already ahead. And if you’re not, well, now you know why your AI-generated content isn’t getting the visibility you hoped for.
Why human expertise still matters

The elements AI can’t replicate (yet)
Despite all the advancements in generative AI, there are still important elements that AI simply can’t give you:
- Genuine personal experience: AI can’t tell real stories from your business journey.
- Original insights: AI aggregates and reformulates existing information—it doesn’t create truly novel perspectives.
- Emotional nuance: AI struggles to understand the subtle emotional contexts that make content resonate.
- Cultural relevance: AI often misses the cultural cues and current contexts that make content timely and relevant.
- Authentic voice: AI can mimic style but can’t create a genuinely authentic brand voice.
As I always say, “Great writers will never be replaced by generative AI. Only mediocre ones will.”
Where AI falls short in content creation
Don't get me wrong—AI tools are impressively powerful. But they have fundamental limitations when it comes to creating truly valuable content.
AI doesn’t truly understand your audience's pain points—it only understands patterns in data. It can’t empathise with your customers’ frustrations or anticipate their unspoken objections.
More critically, AI doesn’t know what makes your approach special. It can’t articulate your unique methodology or explain why your perspective matters. These are exactly the differentiators that make your content worth consuming.
The future content differentiator: human touch
As AI-generated content becomes more widespread, the human touch will become the primary differentiator in the content marketplace.
Think about the content creators you personally follow. I bet it's not because they give you the most comprehensive information (which AI can now do), but because of how they provide it—their unique lens, their authentic voice, their particular way of explaining complex topics.
That human element is becoming more valuable, not less, in an AI-dominated world.
How to use AI as a tool, not a replacement
AI for ideation, humans for real insights
Here's where AI truly shines: accelerating the idea generation and research process.
I use AI to:
- Brainstorm topic angles I might not have considered
- Summarise research papers and long-form content
- Identify patterns and gaps in existing content
- Generate potential headlines and hooks
The key is using AI to expand your thinking, not replace it. I always ask AI to generate ideas that I then filter through my own expertise and understanding of my audience.
AI for outlining, humans for editing
Another sweet spot for AI is helping create content structures. I often use AI to:
- Draft initial outlines based on my topic
- Suggest potential sections and subsections
- Identify key points that should be covered
But here's the crucial part: I never accept these outlines as-is. Instead, I heavily edit them to align with my own thought process and to make sure they have a logical flow that makes sense for my specific audience and purpose.
“Make sure your outline makes sense. It's always easier if you had to write it by hand first and you’ve had to do a few briefs, a few outlines. If you try to become, an expert content marketer tomorrow, but use only AI, it’s not going to work because you’re not going to know exactly what good looks like.”
AI for scaling, humans for quality
The smartest approach is to use AI to handle the repeatable, systematic aspects of content creation while reserving human effort for the elements that add distinct value.
For instance, use AI to:
- Create first drafts of product descriptions or how-to guides
- Generate meta descriptions for SEO
- Reformat content for different platforms
Then have humans focus on:
- Adding unique insights and perspectives
- Incorporating brand voice and personality
- Ensuring emotional resonance and cultural relevance
- Fact-checking and adding expert credibility
Scaling content creation with AI across platforms
Once you’ve mastered using AI for individual content pieces, you can use it for efficient cross-platform content. The key is starting with substantial source material and being very specific about what you need.
Let me share a real example: I've been handling marketing for a theatre production, and I needed to create months of daily social content across multiple platforms. Rather than reading the entire script myself and manually extracting post ideas, I fed the full script to Claude with very specific instructions.
My prompt detailed:
- The platforms we were targeting
- The content formats we needed
- Our specific goals for each content type
- The style requirements
- Exactly what type of content to extract from the script
The result? In about two minutes, Claude generated enough solid content ideas for two months of daily posts across multiple platforms. From there, I spent maybe an hour planning, making small adjustments, and briefing our designer on visuals.
The entire process—from script to two months of scheduled content—took just a few hours, when it would have previously taken days.
This approach works for repurposing any substantial content:
- Turn webinar transcripts into a month of social posts
- Transform a detailed case study into multiple blog articles
- Convert a research report into educational email sequences
The secret isn’t just feeding content to AI though—it’s giving extremely specific guidance about what you want extracted and how it should be formatted. When you do this correctly, you’re not replacing human creativity—you’re amplifying it by removing the tedious extraction work that slows us down.
My 5-step process for creating human-centred content
Over the past year, I’ve been using a hybrid approach that incorporates AI without sacrificing the human element. Here’s my exact process:

Step 1: Start with a proper content brief
Unlike many creators who jump straight to prompting AI, I always begin with a detailed content brief—even when I'm writing the content myself.
A good brief includes:
- Working title
- Target audience specifics and their pain points
- Keywords for SEO, and search intent
- Primary call to action
- Your point of view on the topic
- Expert insights
This brief becomes your North Star, making sure that any AI assistance stays aligned with your strategic objectives.
What I still include in my own brief, but don’t share with AI since these are mostly for you:
- Clear performance goals and what you want readers to do next
- Distribution channels and repurposing strategy
Step 2: Use AI to get initial ideas and organise your research
Once I have my brief, I'll use AI to kickstart my research process:
- Get a few potential angles on the topic
- Find a few recent stats and trends worth exploring
- List common questions people have about the topic
The goal here isn’t to get content I can use directly but to expand my thinking and make sure I’m not missing obvious points.
The key here is to create a research bank for the AI to pull the right information from for this particular article. Some things to include in your research bank are:
- Interview transcripts
- Training workshop notes
- Original research reports
- Customer survey data—remember to remove any personally identifiable information ;)
- Multimedia content related to your topic, like podcasts, infographics, webinars, and slide decks
Step 3: Create an outline with clear heading architecture
Next, I’ll create a proper outline with a clear heading hierarchy:
- H1: Main topic (only one per article)
- H2: Major sections
- H3: Subsections
- H4: Minor points (if needed)
I might ask AI to suggest a potential outline, but I always modify it based on my own knowledge and the specific narrative I want to build. This is where your expertise really matters—creating a logical flow that builds your argument effectively.
Once you’re happy with your outline, get AI to create your ‘ugly first draft’. But to save yourself the pain of editing crappy content, make sure to include your personal voice & tone style guide.
Formatting content for maximum impact in AI search
When it comes to structure and formatting, AI search has specific preferences that differ from traditional search. And although AI search might favour the narrative, story-driven format of typical news articles for context. It also prioritises content that’s structured for usefulness and action.
If you're creating actionable content (which you should be), make sure you include steps and lists. Don’t just describe the process—break it down into clear, sequential actions that someone can actually follow. It’s the difference between “improving your LinkedIn presence requires consistent posting” and “Step 1: Schedule 30 minutes every Monday to draft your weekly LinkedIn content.” And so on.
The engineering mindset wins here. Think about how developers document processes: clear, concise, structured, and without unnecessary flourish. That’s the kind of content that performs best in AI search.
Other formatting elements that make a difference:
- Clear subheadings that answer specific questions
- Bulleted or numbered lists for easy scanning
- Bold text for key takeaways
- Tables for comparative information
- Visual elements that enhance understanding (not just decoration)
Remember, we’re not writing poetry here. The goal is clarity and utility, and that’s exactly what AI search engines are designed to prioritize.
Step 4: Add the human elements AI just doesn’t get
This is the step most people skip, but it’s the most important. Before publishing content, I add:
- Personal stories and experiences
- Original insights from my work
- Client examples (anonymised if needed)
- Controversial or distinctive viewpoints
- Industry-specific context
I literally put these in my outline as [INSERT STORY ABOUT XYZ] to make sure I don’t forget them when later.
Step 5: Edit ruthlessly for voice, authenticity, and accuracy
Finally, whether I’ve drafted the content myself or used AI to help with sections, I edit everything with a ruthless focus on voice, authenticity, and accuracy:
- Replace generic phrases with more specific, engaging language
- Cut any paragraph that doesn’t add distinct value
- Read everything aloud to catch unnatural phrasings
- Add conversational elements like asides and rhetorical questions
- Make sure every section passes the “could only I have written this?” test
This editing stage is where mediocre content becomes great content. Don’t skip it.
Real-world examples: before and after
Example 1: blog article intro
AI-only version
”Content marketing is important for businesses seeking to establish thought leadership. By creating valuable content, companies can attract potential customers and nurture relationships. It is crucial to maintain consistency in publishing schedules and focus on quality over quantity.”
Human-edited
“Let’s be real: everyone and their neighbour is doing ‘content marketing’ these days. But there’s a massive gap between pumping out blog posts to feed the Google gods and creating the kind of content that makes readers stop scrolling and think, ‘Damn, these people get me.’ The former might check a box on your marketing to-do list—the latter builds an audience that actually cares what you have to say next.”
See the difference? The second version has a point of view, a distinct voice, and it challenges the reader rather than stating bland truths.
Example 2: email follow-up
AI-only version
“Thank you for downloading our ebook. We hope you find the information valuable. In the coming days, we will send you additional resources to help you implement these strategies. Please let us know if you have any questions.”
Human-edited
“That monster-sized PDF you just downloaded? I created it after watching three clients make the exact same mistake that cost them each over $10K in wasted ad spend. (Ouch.) Before you dive in, quick question: which of the three challenges mentioned on page 7 is keeping you up at night right now? Send me a reply and let me know—I’ve got some additional resources that might help with that specific issue.”
The human version creates immediate engagement, demonstrates expertise, and invites conversation rather than just delivering information.
Example 3: LinkedIn post
AI-only version
“Leadership in challenging times requires resilience and adaptability. Research shows that effective leaders maintain open communication with their teams and prioritize employee well-being. By fostering a supportive environment, organisations can navigate uncertainty more successfully. What leadership strategies have you found most effective during difficult periods? #leadership #businessstrategy #management”
Human-edited
“Just spent 3 hours in a crisis meeting where our biggest client threatened to walk. My stomach was in knots.
Instead of hiding it, I told my team exactly what happened. No sugar-coating.
Then something amazing happened: Two junior team members proposed a solution none of us senior folks had considered. We implemented it this morning.
Client just called: ‘This is why we work with you.’
Lesson: The best leadership strategy isn’t pretending you have all the answers. It’s creating enough psychological safety that your team will save your ass when you don’t.
Anyone else learn this lesson the hard way?”
See the difference? The first version offers generic advice with no real perspective. The second shares a specific, vulnerable moment that hooks the reader emotionally, demonstrates authentic leadership through storytelling, and invites genuine conversation rather than performative engagement. The human version feels like it came from a real person with real experiences—because it did.
Action plan: Bringing the human touch back to your Content
Quick assessment: How AI does your content sound?
Ask yourself these questions about your recent content:
- Could this content have been written by any of your competitors?
- Does it contain specific examples from your experience?
- Would someone who knows you recognise your voice in this content?
- Does it take a clear position, or does it try to cover all possible angles?
- Does it contain insights that couldn’t be easily found elsewhere?
If you answered “yes” to the first question or “no” to the others, your content probably reads like it was AI-generated – even if it wasn’t.
Practical steps to infuse personality into your content
1. Create a distinctive point of view document
Start with 3-5 strong opinions you hold about your industry that others might disagree with. Add to your doc as you come up with specific point of views on specific topics later on.
2. Develop a stories bank
Keep a database of personal anecdotes, client scenarios, and failures/successes that you can draw from. I build mine in Notion and organise it just like any other content bank so it’s easy to pull from when needed.
3. Record yourself (or your subject matter experts) speaking
Sometimes it’s easier to capture your natural voice by speaking, then transcribing and editing, rather than writing from scratch. And those of us who started in marketing before AI know, this step used to take ages, so don’t skip it now that it’s so much faster to do.
4. Add “only I could write this” elements
For every piece of content, include at least one insight, story, or perspective that could only come from you or your company.
5. Get feedback from people who know you
Ask colleagues or friends: “Does this sound like me?” Their feedback will help you identify where your authentic voice is missing.
Tools to maintain quality
To help you implement this approach consistently, use these :
- A content brief template (like the one shared in my workshop)
- An AI enhancement checklist for reviewing AI-assisted content, like the steps in this article
- A voice & tone verification process for maintaining authenticity, like your personal style guide
Remember: tools and templates are meant to enhance your creativity, not replace it. Use them as guardrails, not crutches.
We need human content more than ever
The irony isn’t lost on me: I’m writing an article about avoiding AI-generated content at a time when AI is becoming more sophisticated by the day.
But that’s exactly why the human touch matters more now, not less. As AI content becomes more common, truly human content will stand out like a lighthouse in a sea of mediocrity.
Your readers don’t want more content. They want better content. They want content that feels like it was created specifically for them by someone who genuinely understands their challenges. They want content that doesn’t just inform, but resonates.
And no AI—regardless of how many parameters it has—can create that resonance without the human touch.
So go ahead, use AI as a tool in your content creation process. I certainly do. But never forget that your unique expertise, voice, and perspective are what truly connect with your audience.
That’s something AI can’t fake—and it’s your greatest competitive advantage in a world drowning in generic content.
Ready to bring the human touch back to your content? Start by reviewing your most recent piece with the assessment questions above, then implement at least one of the practical steps I’ve outlined. And if you're still struggling, hit me up—helping people find their authentic voice is literally one of my favorite thing to do.
About the author
Sophie Michaud is a content strategy advisor and Herizon alumna who helps tech startups cut through the noise with content that actually converts. Specialising in B2B SaaS and analytics, Sophie brings a refreshingly data-driven approach to content strategy that has helped companies grow organic traffic and generate pipeline revenue. She’s also the co-host of the podcast ‘like, is this legit?’.
This article was created with the help of AI during a live workshop with the Herizon community, where Sophie demonstrated her process for using GenAI to create high-impact content that doesn’t sound like it was written by a robot. What you just read was the actual live demo on how to use GenAI for content creation—showing exactly how to combine AI efficiency with that crucial human touch. And yes, it was edited by a human.
When she’s not leading multilingual content operations or crafting multimedia content strategies, Sophie advises tech startups on building content engines that drive measurable business results. The workshop materials and transcripts
used to generate this draft came from several training sessions Sophie hosted for the Herizon community.
Need help making your content suck less and more human? Sophie offers marketing consulting services for tech startups looking to stand out in crowded markets. Reach out at info@sophiemichaud.com to discuss how she can help transform your content strategy from generic to genuinely effective.