You’re assuming that the detector can be trusted. The detector could be someone who is being paid to mislead you on purpose.
If AI presence really matters to you, you need to trust your own two eyes for this sort of thing. Offloading that work to someone else is a considerable risk.
And to be honest, detecting AI is pretty fast once you’re able to spot it. I can spot the typical variants of AI art in just a few seconds.
You need to train your ability to spot AI.
AI art has a very distinctive style. Weird shadows, impossible architecture, and having a blatantly incorrect number of fingers are dead giveaways.
AI text tends to talk at you rather than with you. It has difficulty remembering context, so it tends to forget what you said 10 lines ago.
How about we figure out a better way to detect and sort it so I don’t have to waste time making judgements?
You’re assuming that the detector can be trusted. The detector could be someone who is being paid to mislead you on purpose.
If AI presence really matters to you, you need to trust your own two eyes for this sort of thing. Offloading that work to someone else is a considerable risk.
And to be honest, detecting AI is pretty fast once you’re able to spot it. I can spot the typical variants of AI art in just a few seconds.
I’m not assuming anything, I’m saying we should make something happen.
The halting problem makes that somewhat impossible to make. Detectors have a notable weakness with detecting themselves.