My articles VS the AI detector Pangram
I recently started writing on substack and currently have two articles written, where if I use the AI detector built into the platform, it shows around 2-4% AI assisted with the rest human written. I checked the articles of some of my other friends (those whom I assumed didn't use AI assistance) and in their articles the AI detection was pretty high, 60 to 70 percent or more (even 90+) and I had a very glaring question about that particular finding.
Of course, the AI detection is a flawed program conducted by another AI program itself so its easy to confuse if you write in a similar style when compared to that of an AI, so is it just that my writing is at such an amateur stage, riddled with errors, that its easy to detect it's not AI writing or have others used AI assistance to such an extent?
Honestly, I think another question I'm indirectly asking here is that if its good or bad that the AI detection is so low on my articles?
Any insight will be greatly appreciated, so thank you in advance!