Comment on Trains cancelled over fake bridge collapse image
BanMe@lemmy.world 2 months agoI am guessing the reporter wanted to remind people tools exist for this, however the reporter isn’t tech savvy enough to realize ChatGPT isn’t one of them.
9bananas@feddit.org 2 months ago
afaik, there actually aren’t any reliable tools for this.
the highest accuracy rate I’ve seen reported for “AI detectors” is somewhere around 60%; barely better than a random guess…
rockerface@lemmy.cafe 2 months ago
The problem is any AI detector can be used to train AI to fool it, if it’s publicly available
9bananas@feddit.org 2 months ago
exactly!
using a “detector” is how (not all, but a lot of) AIs (LLMs, GenAI) are trained:
have 1 AI that’s a “student”, and 1 that’s a “teacher” and pit them against another until the student fools the teacher nearly 100% of the time. this is what’s usually called “training” an AI.
one can do very funny things with this tech!
for anyone that wants to see this process in action, here’s a great example:
Benn Jorda: Breaking The Creepy AI in Police Cameras
frongt@lemmy.zip 2 months ago
en.wikipedia.org/…/Generative_adversarial_network
SLVRDRGN@lemmy.world 2 months ago
Someone commented a reply which I thought worthy of highlighting:
“I need privacy, not because my actions are questionable, but because your judgement and intentions are.”