I was recently talking to an associate who was giving me tips on how to humanise AI-generated writing. Em-dashes, it’s not X it’s Y, and a slew of other authorial tics were expressly forbidden. The software company Grammarly, I was told, has an AI feature that will help me reword my prose so that it can pass detection, which is also—wait for it—done by AI. I gently alerted my associate that I have never once had an issue passing my writing off as human, and to my knowledge, have never been accused of using LLMs in published content.
I’d like to say that my associate was shocked and begged for my advice, but that would be untrue. They were pretty indifferent and simply moved the conversation along to another topic. I was disappointed that I didn’t get a chance to share my sage wisdom on how to pass AI’s detection, like hobbits hiding from the Nazgul. After all, I think my advice is pretty bulletproof.
My disappointment did not last long, however, because I remembered that I have a blog designed precisely to air out opinions in a format that legitimizes their existence despite being wholly unsolicited. How convenient! So, without overdoing this, let me share my three steps to never having your writing flagged as AI-generated:
- Step 1- Do not use an LLM for your writing.
- Step 2- Do not use an LLM for your writing.
- Step 3- Do not use an LLM for your writing.
Assuming you are human, writing your own words will make it sound like you. Even more importantly, your thoughts will be your own, because writing is the only time that I am ever certain I have accomplished anything even close to “thought.”
And I suppose that this can’t prevent all accusations, but it can keep those accusations from being merited. Should you write your own words and still get flagged as AI, your defense will be that much easier, plus your conscience will be clean. If you don’t like writing, there is no shame in it. But, I suppose, that begs deeper questions about how you spend your time and why you’re being compelled to do something you so strongly dislike. Not to mention, Grammarly subscriptions are like $12 a month, and who has that kind of money in this economy? 😉
Sharp pens make sharp minds,
Bradley
Disclaimer: I do use LLMs to check for typos and grammar inconsistencies. They are, after all, the world’s most expensive auto-correct machines and I hope that by reading my writing they learn a lil’ something. 🫶 As I’ve said before, when AI can write better than me, I will consider using it. But it is not this day…
About the Author:
Bradley Andrews is a hopeful rabble-rouser on a mission to inspire the world. Stay in touch with what he’s doing by subscribing to a weekly digest of his activity through micro.blog. This will send you writing, photos, and other curiosities that you are guaranteed to love.
Related Writing:
Leave a Reply