What if everything you read on the internet was produced using Large Language Models? That all content you thought you are reading was generated by an AI? Would you be able to tell the difference? Can you read something and say, ‘no human would right this?’
I have some favorite blog authors as well as podcasters. I read and listen to them regularly and wondered how they can be so productive. How they can churn out content as often as they do. I also looked back at my own writing and wondered how original I am? And if you do not know me well enough those thoughts drove me into a black hole of trying to see if how to determine if content is real, or LLM generated.
First, I found a paper that describes multiple ways to determine if content was LLM generated. Discovered another article that shares words that tip you off that something was written by and LLM. And finally started kicking around with ZeroGPT as a site that allows you to put content in, and it will give you what percentage was written by an LLM! Of course I took my recent posts, and then incredibly old posts and wondered what it would say about what I wrote.
My most popular post ever was done recently, telling a story about a tree falling in my yard.
It thinks the first two sentences were written by AI GPT, but the rest were not. So, it believes my post was written by a human, which is good. I am not sure a LLM would think of a story like this, nor compare it to a relationship. I went through my team building posts, and then decided to go back to the really old posts.
First, one of my favorites Disney being the new ’Historians’ as Biopics distort history. This one turned out to be 100% human written. Next my quick thought about sometimes you paint the car, sometimes you paint the fish, also 100% human written. Other ones I won’t link also turned out to be estimated at 100% human written. And one, which is a story I have told often and a quote I use often Even a Blind Squirrel eventually finds a Nut (according to Google this is the most clicked on link from their search), was estimated to be 90% human written, and 10% AGI. Strange as that was written in 2017!
Of course, next I started putting some of the people I read, and their content to the ZeroGPT test. Most came out with excellence (5% or lower AGI) but there were a few outliers whose content was marked as GPT. I would not say it shocked me that at least a few people were leveraging LLMs to write their content. What I think was shocking is some posts were marked as almost 100% written by AI.
I am not calling them out by putting any of their content here. And I should note, for the past year or so I have been using a text editor that has some AI built into it. It attempts to correct my grammar, my spelling etc. The AI suggests wording that might be more appropriate. Sometimes I agree, sometimes I do not agree. It is my choice while writing and editing.
A primary concern when using LLM-generated content is the potential lack of authenticity. Bloggers may unintentionally present ideas or viewpoints that are not their own, compromising their credibility and the trust of their audience. Reader’s value genuine voices and personal experiences, which LLMs may struggle to replicate.
Looking back at my posts, what amazed me was the influence of others on my writing. Whether it be Bruce Lee (who I quote a lot) authors like Adam Grant, Simon Sinek, other podcasts like TWIT, Hidden Brain, Huberman labs, or even just friends, my views, thoughts, and ideas are an accumulation of what I pick up. And to quote Bruce again ‘Learning is never cumulative; it is a movement of knowing with no beginning and no end.’ I have regurgitated information like an LLM, but I knowingly make choices on what I keep and what I throw out. The algorithm is not guided by any rules, other than the decisions I make. I will continue to ingest from talented authors and podcasters, and I will use them as a guide to discover more about myself. I just hope my output is not decidedly systematic and that my posts are marked as written by an LLM.
If you made it through the whole post, one paragraph was written by an LLM, wonder if you spotted it. To answer the question, I asked in the beginning. If everything were written by an LLM, it would still be up to us readers to decide if we want to read it, if we stop, then maybe the production of it will be reduced.
This opinion is mine, and mine only, my current or former employers have nothing to do with it. I do not write for any financial gain, I do not take advertising and any product company listed was not done for payment. But if you do like what I write you can donate to the charity I support (with my wife who passed away in 2017) Morgan Stanley’s Children’s Hospital or donate to your favorite charity. I pay to host my site out of my own pocket, my intention is to keep it free. I do read all feedback, I mostly wont post any of them.
This Blog is a labor of love, and was originally going to be a book. With the advent of being able to publish yourself on the web I chose this path. I will write many of these and not worry too much about grammar or spelling (I will try to come back later and fix it) but focus on content. I apologize in advance for my ADD as often topics may flip. I hope one day to turn this into a book and or a podcast, but for now it will remain a blog. AI is not used in this writing other than using the web to find information, and the one paragraph if you can find it. Images are from ZeroGPT, I do not have a financial investment with them, I used the free version FYI.