Should we wait for a writing to be detected as more than 50% using AI while that person has benefited from using AI. I have discussed with some media writer friends and they said that Chat GPT can give us information that cannot be detected by AI detector tools.
You are viewing a single comment's thread from:
For a writing to be passed as an AI the threshold is usually above 90% so anything below that is actually still in the gray and even those detectors can't be trusted.
Article written with AI has some similarity and styles to it unless it is paraphrased.
Yup, the use of AI has a very obvious format and feel especially if you know the person has never written and read a book before, it is clearly suspicious.
As a writer I can feel which writing that comes from that own thoughts, even if that person uses translation, the language that person uses cannot possibly have (padanan kata) good word choice.
Even I who have written in the media feel that writing a feature article is very difficult and for someone who always complains that he/she can't write more than 500 words can suddenly write a feature article with 1000 words? Is it impossible?
Using AI is not a mistake but a deliberate form.
Well that's just silly. It's like complaining that someone wrote an article with Microsoft Word rather than painstakingly scribbled it down with a feather pen.
The only important question is whether the result is good. If the result is bad, it doesn't matter what tool they used or no tool at all. If the result is good, it likewise doesn't matter. Judge the work.
Those are clearly two different contexts. Using AI is just copy past and not implement what is in your mind into writing. It would be ridiculous if this platform allowed anyone to move anything in text form.
Let's simplify the example, someone uses AI to write a motivation letter for a scholarship, the person is accepted for the scholarship. While there are people who really write their motivation letter but they fail to get the scholarship because someone else uses AI.
Is AI's work bad? They are very well programmed to help human work, not to deceive others. On the contrary, no matter how bad a writing is, if it is written honestly, it is still better than an informative writing using AI.
You've never actually tried working with AI, have you? It's not just copy-paste. It's more like riding a horse. You direct, it acts, you refine.
If an AI can write a better motivation letter than you can, then the person who is smart enough to use the AI as an adjunct is more capable than you. It's that simple.
The objection that anything “written honestly” is better than something which uses a tool is just the age-old complaint of painters bitching about cameras and telling us how they aren't art, that they're soulless.
That was bullshit before the turn of the century, and it remains so.