Stop Blaming AI for Your Boring LinkedIn Posts

LinkedIn is drowning in what people are calling "AI slop".

Generic posts. Lifeless comments. Content that reads like it was written by a robot having a particularly uninspired day. And yeah, over 54% of long-form LinkedIn posts are now AI-generated. The backlash is real. Usage of the term "AI slop" increased 9x in 2025 compared to the previous year.

But here's the uncomfortable truth nobody wants to hear: AI didn't create boring content. It just made it easier to produce at scale.

Most corporate LinkedIn posts were already formulaic rubbish before ChatGPT came along. We just couldn't pump them out fast enough to notice the pattern.

The Real Problem Isn't the Tool

I've been experimenting with AI for content creation for the past year. I've set myself the challenge of releasing 48 blog posts and posting daily across LinkedIn, Instagram, and TikTok.

Writing isn't my strong suit. I'm a busy CEO with five hours of meetings most days. I've got four kids. The idea of consistently creating content was impossible a year ago.

Now? I'm reaching over 10,000 business professionals weekly on LinkedIn. My following is growing 1-2% every week. I sold out an AI event. Built a community around 21 Degrees that we weren't reaching before.

And I'm using AI for almost all of it.

The difference between my content and the AI slop flooding your feed? Context. Input. Authenticity.

Why LinkedIn Became Ground Zero for Generic Content

LinkedIn isn't fun like Instagram or TikTok. Posting on LinkedIn is work.

You're representing your company. Your personal brand. There are guardrails. Your voice might not match the company you work for. You're trying to create content at scale whilst keeping it engaging enough to stop people scrolling.

Everything on LinkedIn still runs on the attention economy.

The pressure to consistently produce quality content within those guardrails becomes massive stress. People overthink everything. They need volume but they need it to be good.

So they turn to AI. They blindly use it to write short paragraphs, comment on posts, do more with less.

And the LinkedIn community is calling them out for it.

The Skills Gap Is About Humans, Not AI

Here's what the big foundation model companies are saying: the blocker for AI isn't the AI anymore. It's humans figuring out how to use it. Research backs this up. Fortune 500 executives are worried about a "critical thinking gap," not an AI capability gap.

The technology has caught up. We haven't.

If you go into ChatGPT and ask it to create a post about yoga positions, you'll get generic rubbish that anyone could get. Upload that online and you're giving nobody any insights. Nothing valuable. Nothing interesting.

But if you load in tonnes of your former writing, your video transcripts, anything that teaches the AI how you talk and what you're likely to say? Then add your own personal opinion on top?

You get something totally unique that talks about your opinion in your tone of voice.

What Actually Makes Content Engaging

The data tells us what works. Employee-shared posts outperform brand posts by 5-10x.

Personal profiles get 7-8% engagement. Company pages? 1-2%.

Authenticity wins.

I've had clients tell me they can hear my voice in their head when they read my AI-assisted posts. That's the gold standard. That's what separates valuable content from slop.

It's not about whether AI touched your content. It's about specificity. Genuine experience. Taking a stance. Showing your working, not just your conclusions.

The patterns of forgettable posts are obvious: vague platitudes, humble-bragging, recycled tips, opinion-free observations that could apply to anyone.

How to Actually Use AI Without Creating Slop

I've built a system that works. Here's how I do it:

First, I experiment with different LLMs. ChatGPT, Claude, Gemini. Which one nails your voice? Which one gives you the best results?

Then, I feed it context. Loads of it. I've loaded Claude with huge amounts of my transcripts from talking to camera. My thoughts, my feelings, my conversations. I've

even tied in a couple of my favourite writers—Hunter S. Thompson and Douglas Adams—to give it that quirky wit that matches my ADHD sensibilities.

I create a skill in Claude. This means I can just ask it to write me a post and it understands the context. It knows what I would say, what I wouldn't say, where I like things quirky and where I like them dead on.

Here's the crucial bit: I have conversations with AI, not commands.

If writing is something you struggle with, sit in your car and have a ten-minute conversation with your LLM. Then ask it to craft a post from that information.

Don't accept the first answer. Give feedback. The more it learns about you, the better it gets.

The Practical Steps

Load your transcripts into Claude. All of them. Video transcripts, former writings, posts, anything that shows how you communicate.

Ask it to create a tone of voice document based on your transcripts. Read that document carefully. What has it got right? What has it got wrong? Which direction do you actually want to go?

Use that document to create a skill within Claude. This makes it quicker and easier to use consistently.

That tone of voice document is gold. You can load it into ChatGPT, Gemini, Cursor, any tool you use. The more background information and context you provide, the more authentic your content becomes.

When to Stop Using AI

Here's my advice: if you can't upgrade the context you're feeding into your AI, stop using it.

It's probably doing your reputation more damage than it's helping.

People using AI aren't training their models well enough. The bots they're creating use out-of-the-box software that lacks that definitive tone. That feeling of you.

Upgrade your software. Upgrade the context. Or don't use it at all.

Being Opinionated Is Your Competitive Advantage

In an AI-saturated feed, being opinionated is your edge.

Algorithms can't replicate genuine expertise. They can't replicate hard-won perspective. They can't replicate the specific journey you've been on in your industry.

LinkedIn's algorithm rewards this. Content that triggers meaningful conversations—defined as three or more comment exchanges between different participants—receives 5.2x the amplification of comparable posts without discussion depth.

Posts on emerging topics or with new viewpoints on established topics get 165% more distribution.

Being willing to take a stance pays off.

The Bottom Line

AI isn't the problem. Lazy thinking is the problem.

The quality of your input directly determines the quality of what AI produces. Feed it generic prompts, get generic content. Feed it your authentic thoughts, your specific experiences, your genuine perspective? You get content that sounds like you.

Stop optimising for engagement. Start optimising for impact.

Be useful. Be real. Be willing to alienate the wrong people.

And if the post that comes out the other end doesn't sound like you? Don't use it.

Be you. That's the simplest rule. That's the only rule that matters.