Page 1 of 1
Two Stories
Posted: Fri Jan 26, 2018 3:42 am
by _EAllusion
Here's two interesting stories that combined form a probable Black Mirror hellscape probably coming to us soon:
First we have a story about a college professor who inadvertently ended up being the target of a online white supremacist mob:
https://www.aaup.org/article/new-realit ... mqhK6inGUlIt's a fascinating read about their systematic attempts to destroy his personal life. They leaned heavily on creating fraudulent personas and documents to make false allegations about him to everyone in his orbit.
Then we have a story about how people with hobbyist level technical skills are making AI-generated porn with quite convincing face-swapping on a level what was cutting edge not that long ago.
https://motherboard.vice.com/en_us/arti ... isy-ridley Combine this with the fact that it new A.I. tech has made it possible to say any fed text in a voice that mimics audio samples of a person.
https://www.scientificamerican.com/arti ... any-voice/What you end up with is consumer-level ability to generate completely faked, but superficially compelling audio-visual presentations of people saying whatever being available to the masses relatively soon. This has nightmare level implications for politics, but imagine what the white supremacists described in the first story, or really anyone with a surplus of grudges and a deficit of ethics, are going to do with that.
Re: Two Stories
Posted: Fri Jan 26, 2018 3:47 am
by _DoubtingThomas
EAllusion wrote:What you end up with is consumer-level ability to generate completely faked, but superficially compelling audio-visual presentations of people saying whatever being available to the masses relatively soon. This has nightmare level implications for politics, but imagine what the white supremacists, or really anyone with a surplus of grudges and a deficit of ethics, described in the first story are going to do with that.
I am a big believer in A.I., I think it is going to do more good than harm, I am not so worried about fake audio-visuals. I see A.I. as one of our last hopes, we desperately need A.I..
Re: Two Stories
Posted: Fri Jan 26, 2018 4:04 am
by _EAllusion
DoubtingThomas wrote:EAllusion wrote:What you end up with is consumer-level ability to generate completely faked, but superficially compelling audio-visual presentations of people saying whatever being available to the masses relatively soon. This has nightmare level implications for politics, but imagine what the white supremacists, or really anyone with a surplus of grudges and a deficit of ethics, described in the first story are going to do with that.
I am a big believer in A.I., I think it is going to do more good than harm, I am not so worried about fake audio-visuals. I see A.I. as one of our last hopes, we desperately need A.I..
Think about how many people fell for transparently fake news stories last election cycle. For example, remember the fake news story about the Pope endorsing Donald Trump for president? That was laughable, but there is plenty of reason to think it held some influence. Imagine if instead of it just being a social media equivalent of a chain-mail with faked signifiers of journalism, it was attached to a believable video of the Pope endorsing Donald Trump. We already had a situation where a lie traveled much further than its correction, but that would probably exacerbate the problem a great deal more. Now imagine a situation where people are creating fake videos of candidates saying things they did not to attack them. Someone could simply fake Donald Trump using explicit racial epithets to discredit him or to false flag discredit the people who would discredit him.
Sounds bad, right? What's really bad though is that actual evidence of actual wrongdoing will be dismissed as fake by anyone who wants to believe whatever they want about their leaders. Once you are in the wild west of where video and audio can and is being faked, counterfaked, and false flagged by enemies, it becomes easy to rationalize any bad information as fake.
Ok, it's already getting bad, but that's scratching the surface. Police are allowed to lie to suspects in interrogation in order to advance their investigations. What if police were allowed to use faked videos to coax confessions? That doesn't seem too far fetched.
What if people agitating for a genocide in a country use faked videos validating the case they are using to foment hated?
There's a lot of bad possibilities with this. The one that I started the thread about was how this kind of technology could be used by online mobs like the one described in the first story to be more effective at disrupting people's lives.
Re: Two Stories
Posted: Fri Jan 26, 2018 4:22 am
by _DoubtingThomas
EAllusion wrote:Think about how many people fell for transparently fake news stories last election cycle. For example, remember the fake news story about the Pope endorsing Donald Trump for president? That was laughable, but there is plenty of reason to think it held some influence. Imagine if instead of it just being a social media equivalent of a chain-mail with faked signifiers of journalism, it was attached to a believable video of the Pope endorsing Donald Trump. We already had a situation where a lie traveled much further than its correction, but that would probably exacerbate the problem a great deal more. Now imagine a situation where people are creating fake videos of candidates saying things they did not to attack them. Someone could simply fake Donald Trump using explicit racial epithets to discredit him or to false flag discredit the people who would discredit him.
Sounds bad, right? What's really bad though is that actual evidence of actual wrongdoing will be dismissed as fake by anyone who wants to believe whatever they want about their leaders. Once you are in the wild west of where video and audio can and is being faked, counterfaked, and false flagged by enemies, it becomes easy to rationalize any bad information as fake.
Ok, it's already getting bad, but that's scratching the surface. Police are allowed to lie to suspects in interrogation in order to advance their investigations. What if police were allowed to use faked videos to coax confessions? That doesn't seem too far fetched.
What if people agitating for a genocide in a country use faked videos validating the case they are using to foment hated?
There's a lot of bad possibilities with this. The one that I started the thread about was how this kind of technology could be used by online mobs like the one described in the first story to be more effective at disrupting people's lives.
Like you describe fake news is already a big problem, the idiots are going to fall for the dumbest Facebook s*** anyway. So I you understand your concerns, but at the same time I believe A.I. can fix the fake news problem.
Re: Two Stories
Posted: Fri Jan 26, 2018 9:25 am
by _subgenius
DoubtingThomas wrote:..., we desperately need A.I..
Why? We haven't really "needed" it at all to date.
Re: Two Stories
Posted: Fri Jan 26, 2018 9:28 am
by _subgenius
EAllusion wrote:Here's two interesting stories that combined form a probable Black Mirror hellscape probably coming to us soon:
(Emphasis mine)
And on another point, One should often consider the difference between imagination and prediction before setting one's hair one fire.
Re: Two Stories
Posted: Fri Jan 26, 2018 2:05 pm
by _EAllusion
subgenius wrote:EAllusion wrote:Here's two interesting stories that combined form a probable Black Mirror hellscape probably coming to us soon:
(Emphasis mine)
And on another point, One should often consider the difference between imagination and prediction before setting one's hair one fire.
Black Mirror already did an episode on face-swapping. This is exactly the sort of technology they would appropriate into a story.
I can't help it if you can't see clear applications of technology. I'm proposing that people who photoshop fake screenshots to defame a professor might make fake videos when that technology becomes available to them. Quite a leap.
Re: Two Stories
Posted: Fri Jan 26, 2018 7:48 pm
by _DoubtingThomas
subgenius wrote:DoubtingThomas wrote:..., we desperately need A.I..
Why? We haven't really "needed" it at all to date.
How do you know? and how do you know it can't help us cure diseases and solve many of our problems? But I understand you don't need A.I. subgenius, all you need is God and religion.
Re: Two Stories
Posted: Sat Jan 27, 2018 12:08 am
by _canpakes
DoubtingThomas wrote:But I understand you don't need A.I. subgenius, all you need is God and religion.
Not true. He also needs to compulsively rely on and overuse a limited number of tired and boring figures of speech in order to attempt to communicate - such as, "butthurt", or "hair on fire".