The third one could easily be satire. Imagine that a politician is accused of stealing from the public purse, and issues a meme-worthy press statement denying it, and someone generates AI content of that politician claiming not to have stolen a car or something using a similar script.
Valid satire, fair use of the original content: parody is considered transformative. But it should be labeled as AI generated, or it's going to escape onto social media and cause havoc.
It might anyway, obviously. But that isn't a good reason to ban free expression here imho.
Respectfully disagree. Satire should not be labelled as satire. Onus is on the reader to be awake and thinking critically—not for the entire planet to be made into a safe space for the unthinking.
It was never historically the case that satire was expected to be labelled, or instantly recognized by anyone who stumbled across it. Satire is rude. It's meant to mock people—it is intended to muddle and provoke confused reactions. That's free expression nonetheless!
So when we have perfect deep fakes that are indistinguishable from real videos and people are using it for satire, people shouldn’t be required to inform people of that?
How is one to figure out what is real and what is a satire? Times and technologies change. What was once reasonable won’t always be.
- "How is one to figure out what is real and what is a satire?"
Context, source, tone of speech, and reasonability.
- "Times and technologies change."
And so do people! We adapt to times and technology; we don't need to be insulated from them. The only response needed to a new type of artificial medium, is, that people learn to be marginally more skeptical about that medium.
Nah. Satire was always safe when it's not pretending to have documented evidence of the thing actually happening.
Two recent headlines:
* Biden Urges Americans Not To Let Dangerous Online Rhetoric Humanize Palestinians [1]
* Trump says he would encourage Russia to attack Nato allies who pay too little [2]
Do you really think, if you jumped back a few years, you could have known which was satire and which wasn't?
The fact that we have video evidence of the second is (part) of how we know it's true. Sure, we could also trust the reporters who were there, but that doesn't lend itself to immediate verification by someone who sees the headline on their Facebook feed.
If the first had an accompanying AI video, do you think it would be believed by some people who are willing to believe the worst of Biden? Sure, especially in a timeline where the second headline is true.
Valid satire, fair use of the original content: parody is considered transformative. But it should be labeled as AI generated, or it's going to escape onto social media and cause havoc.
It might anyway, obviously. But that isn't a good reason to ban free expression here imho.