Stop using LLMs to write your posts because everyone can tell
Using AI to write your social media posts destroys your credibility and makes readers question if your ideas are even real.
Business · Source: Hacker News
What happened
Bryan Cantrill published a sharp critique of AI generated content on LinkedIn. He argues that using large language models to write posts is like walking around with your intellectual fly open. Everyone sees the embarrassing mistake but nobody points it out. Social platforms actively push users to rewrite their thoughts with AI, which makes the problem significantly worse.
The writing produced by these models is stylistically grating and highly predictable. It relies heavily on single sentence paragraphs, specific em dash constructions, and predictable phrasing like not just but also. Readers easily spot these tells because the demographic of people who recognize AI text is rapidly expanding. Once they spot the artificial style, they immediately stop reading your content.
Cantrill notes that models remain excellent for brainstorming, reading comprehension, and editing. You can safely ignore an AI editor without starting a fight over grammar rules. But when you let the model do the actual writing, you sacrifice your unique voice and replace it with generic filler. Readers are left wondering if your actual content is real or just generated fan fiction.
Key facts
- November 11, 2025 — Date the critique was originally posted on LinkedIn
- December 5, 2025 — Date the post was published on The Observation Deck
Why it matters
Founders building in public need trust more than they need content volume. When you use an AI to write your updates, you signal that you do not care enough to speak directly to your audience. Readers start wondering if your product updates, revenue metrics, and technical claims are also generated fiction. Your audience wants your actual perspective, flaws and all. If you hide behind a language model, you destroy the exact credibility you are trying to build.
Authenticity is rapidly becoming a premium asset in the business world. As the internet floods with synthetic text, a genuine human voice stands out more than ever. People who write their own content will capture the high value attention from investors and customers. Those who automate their writing will be silently ignored by the exact peers they want to reach. The cost of looking like a robot is losing your most valuable professional relationships.
For builders
Use AI for editing but not writing
Draft your posts yourself to keep your authentic voice intact. Use models strictly to check your logic or suggest structural improvements after the fact. You retain the final say without sounding like a generic robot to your audience.
Stop using predictable AI formatting
Avoid the standard AI output style at all costs. Drop the excessive emojis, the predictable transition phrases, and the endless single sentence paragraphs. Readers are highly trained to skip this format, meaning your target audience will never see your message.
Trust your own voice over polish
Your audience wants your real thoughts and unique perspective. A slightly clumsy post written by a real founder always beats a perfectly polished AI essay. Authenticity wins the attention economy, so have the confidence to publish your own unvarnished words.
My take
I see founders doing this every single day and it instantly kills my interest in their product. If you cannot be bothered to write your own thoughts, I cannot be bothered to read them. Stop outsourcing your personality to a language model and start respecting your audience.
Original reporting: Hacker News. This is my rewrite and opinion.