Short YouTube demo by LARRUZ showing the GPT-2 Output Detector. The speaker pastes AI-generated text into the detector, which reports a 99.97% AI score, then alters the prompt to request a tone "not typical to AI" and demonstrates that this simple prompt engineering causes the detector to fail to flag the text as AI-generated.
Plagiarism is kind of the main issue when it comes to AI generated text. But now we get to the AI plagiarism checker. This one is called GPT2 output detector demos and you can find it under this URL. So if I copy paste the text here, all I have to do is wait for a bit and it's going to predict how much of this was written by AI. And right away we have it. 99.97% fake. But let's see if we can get around this. Can we fool this with some creative prompt engineering? Let's just try a basic. Now write it in a tone that is not typical to AI. Alright, so now let's copy paste this new text. Oh my God, look at that. By simply telling it right in a tone that is not typical to AI, I pretty much got around this version of the detector.
