Soldato
- Joined
- 17 Jun 2010
- Posts
- 12,495
- Location
- London
Is anyone else genuinely worried about the potential for harm this sort of stuff has? Given Twitter tit-for-tats have caused diplomatic incidents, how long before a deep fake ends up starting a war or some sort of physical skirmish?
The speed at which these are improving as well is crazy (not just video but things like voice imitation). First it was sad cases making porn with their favourite celebrity’s head replacing the original performer, now you have those face-swap apps anyone can use to get a laugh out of... are we going to end up in a deranged world where we won’t be able to tell if the person we’re looking at or listening to is really them?
Machine-learning is cray-cray.