For this week’s blog post, I asked OpenAI’s chatbot called ChatGPT to create a short blog post. The result I got was decently written and could pass as authentic writing, which immediately made me wonder if I’ve read any AI created content online and didn’t recognize it.
The prompt I used was, “in 250 to 300 words, write about the benefits of self myofascial release techniques”.
I chose this subject because self-myofascial release is a modality I’m certified to discuss through NASM’s Corrective Exercise Specialist program. Generally, the AI’s generation was accurate but felt repetitive, like it was rewording the same information to fill up the word requirement. I had to make quite a few edits for the writing to align with my voice and what I felt was important to elaborate on.
In this Google Doc you can see what corrections I made on the second page. For the sake of this experiment, I didn’t want to edit or add entire paragraphs of additional facts and data, I wanted to keep the backbone of what the AI chatbot created. If it were up to me, I would have expanded on what myofascia is and how to use the tools with proper application, but I only asked the chatbot to write about the benefits so I stuck with that.
I think using a chatbot can be helpful in starting a brainstorming process, inspiring ideas, or creating an outline, but users must act with caution. If using a chatbot to write for professional or educational content, it’s important to be authentic and use your own words, voice, and sources. ChatGPT doesn’t cite any of its sources, so it shouldn’t be considered as credible or trustworthy source to get information from.
Leave a Reply