I’ve been a little late to jump on the ChatGPT train.
I first started using it to write business social media posts, usually pasting in what I wrote, asking Chat to “make it sound a little more professional”. It does its work, and I always think how much better it sounds! I find it also useful for writing a letter or a note in a card, taking out my habit of sounding a little pompous. It always succeeds.
More recently, I’ve been using it to identify something in old pictures, like my uncle’s Polish Legion uniform, or information on how an ancestor would have made their way from Ellis Island to their sister’s home in Massachusetts.
Yesterday, I asked it about some wording I pulled from a palm reading I had in 1987. Yes, 39 years ago and I still have the transcription from the reading. It gave me some insightful information.
Now, I realize this is artificial intelligence, so I asked it how it determines what to reply to me. It says it analyzes the words I use, the tone, and whether I’m asking for facts, advice, emotional support, strategy, or creativity. It considers my past context and remembers general things about me to respond better such as, I run a remodeling business, my age, that I work out in the mornings, and I’m exploring my Polish family history. It predicts from that, what would be most helpful using probability and language patterns. And lastly, it adjusts to my style and doesn’t respond to me like it would someone “firing off one-line tech questions.”
From there, it asked me when I ask this question, what am I wondering about. I know there have been many instances of ChatGPT being wrong, or sadly, promoting or assisting in suicide. I am wondering how it “knows” me and how their responses are shaped because I found its response to my palm reading session so long ago very thoughtful, for artificial intelligence.
I’ll continue to use it, with caution, and not get caught up in the idea that it really “knows” me.
Do you use some form of artificial intelligence for information?