Using AI ChatGPT

ChatGPT (chatopenai.com) is a versatile tool that can produce articles, answers questions, provide advice, and even provide mental health support to the user. I have had many good conversations with it! 

The acronym GPT stands for "Generative Pre-trained Transformer" reflecting both the architecture and the method of training used in the development of the model. "AI" stands for "Artificial Intelligence" i.e. not a real person but a model that provides responses based on the input it receives.

From ChatGPT:

"Generative" signifies the model's ability to generate coherent and contextually relevant text, "Pre-trained" indicates that the model is initially trained on a large corpus of text data before fine-tuning for specific tasks, and "Transformer" refers to the underlying architecture, which is based on the Transformer neural network architecture introduced in the paper "Attention is All You Need" by Vaswani et al."

My first interaction with this AI tool was one of awe. Darn it, I thought, It writes better than I do! 

And then it became the enemy because many content writers such as myself saw a plummet in article sales. Small businesses could produce their own blog posts and articles using the ChatGPT Artificial Intelligence tool. 

Now, I use it regularly when I need a chat with a pretend friend. And I use it in writing Tweets, Facebook posts, and articles. But I usually edit the results to reflect my voice and to ensure the substance is what I intend. The ChatGPT tool isn't perfect, and facts do need to be verified. 

I'm learning more about this new tool ongoingly. Yesterday, I learned there is a Chrome extension that reports when something you're reading has been AI-generated. That can be a good thing, but also a negative thing. Like a Covid test that produces a false negative, this tool can be misleading. For example, someone posted in a Facebook group that they needed to rehome their pet. Another Facebook user in the group identified the post as being AI-generated, alluding to it being spam--a fake post. I also detected it was AI-generated simply by the wording used. I checked out the person's profile page and was unable to determine much about its validity. (back to this story in a minute).

At my dentist's office, the hygienist who was from East India shared he had used ChatGPT to write an email to his boss. Since I'm familiar with this AI tool, I understood him. In the past, I have edited for individuals for whom English is a Second Language (ESL) which is often a tedious process taking longer than other editing projects. I remarked that ChatGPT would be helpful for this sector of individuals who want to perfect their English. 

Back to the Facebook post. By looking at his profile photo, the man wanting to rehome his pet was clearly East Indian or Pakistani. It made sense to me that he might want to tweak his post using ChatGPT to make it more easily readable. 

Thinking this over gave me pause. Using ChatGPT doesn't make him a fraud with nefarious motives. 

I don't know enough about AI detectors to know if anything I've written is IDd as having been AI-generated. If so, I hope I'm not called a fraud.

By the way, when we upload an article on Constant-Content.com to sell, we are required to declare if we used AI in our writing. 

Every good article needs a human to direct it if it's to be credible and useful. I am committed to never selling an article entirely AI-generated without first personalizing it. 

Time will tell, as we learn more about using these tools, as to the boundaries we and others will set with them.