Re:Stopping deepfake news with an AI algorithm that can tell...
by hellofunk on Sunday August 09, 2020 @10:40AM
(#24099916)
> You're missing a sarcasm mark.
I don’t see any sarcasm there.
I don’t see any sarcasm there.
But seriously: at best it's going to be a big "maybe, it depends". These services bounce between PR disasters of "complicit purveyors of propaganda" and "oppressive algorithms destroying free speech", so you can't rely on them to be on either side.
As deepfakes get better, the algorithms will have worse detection rate, so it'll end with Zuck testifying to congress that Facebook doesn't want to decide what is real, so everyone can post deepfakes all day long.