Can you tell just by reading whether a post or response was written using ChatGPT? What are the clues?
Can be generic and non specific, and also the human touch may be missing, such as lacking unique personal experiences
Nowadays I am usually spot on in guessing if a post is composed by chathpt. The language lacks human flaws and a certain character that human writing intrinsically has. Miss the good old days when people used their brains and creativity to compose original works.
Yes I spot the chatgpt responses pretty easily, they are pretty generic and no clear viewpoint. The sentences are quite rambling, but doesnt convey a clear opinion or even a clear fact. What can be conveyed in a single line or phrase is explained in 3-4 sentences... This example is not from chatgpt, but demonstrates the rambling - I actually read such a news article a couple of months back in an english newspaper!!! "Today due to rain the school will be closed. The entire year the school was open and the students are very happy that today's rain has caused the school to be closed. There was a lot of rain today, the roads are overflowing and it will be quite difficult for the students to reach the school, especially those who stay far away will find it time consuming to reach school. The teachers also are not required to come to school today........" No idea which city, which school !!! No details of specific reasons (a lot of rain is not enough reason), no clear data or numbers given (how many cms of rain, is it much higher than average etc), no info on which particular road is closed.... it went on like this for 10 more sentences without any details! I notice this especially when I ask to differentiate some 2 different concepts - the distinguishing features is written in the first line, after that, they just rewrite the same point repeatedly using a different set of words.... Recently I had asked one intern to write a research article, I told her to ask chatgpt - she did that, took out the sub-topics and used the structure of the article but wrote her own sentences!
I can easily figure out if its AI generated post. It's not spontaneous. repetitive sentence with a set pattern.
Thanks for the responses, girls. The question was about posts and responses in IL. I have started to see posts from members that are totally different from their earlier style and so got curious. To answer the thread's question, the clues for me are the overuse of commas, longer sentences, and some rare punctuation marks like the double hyphen —. The biggest clue is that each sentence conveys more than a small, simple thought.
Are people actually able to do that? Here i am trying to get one point across and having to write an essay for that
I have a browser extension that auto corrects, includes punctuations and even saves drafts. Including personal experiences is definitely a clue. I think bots don't do that.
Yes, it is easy to make out if a post is AI generated especially when one has interacted with the 'poster' before. I wonder if we will have a cost-effective, user-friendly AI aid that will assist people to make impressive speeches, in the future. Someone who cannot make coherent sentences in any language will then make impactful articulations on any topic under the sun.