They stole my voice with AI • Hacker News Board

Re: They stole my voice with AI

#191

Post by photonthug » Sun, Sep 22, 2024, 10:00 AM UTC

kossTKR wrote: post #153

godelski wrote: post #140

Earlier quoted context omitted.

> what if AI was used to imitate a person saying something blasphemeous? I've been contemplating writing an open letter to Dang to nuke my account. Because at this time you can likely deanonymize any user with a fair amount of comments. As long as you can correlate. You can certainly steal their language, even if not 100% accurate. It may be caution, but it isn't certain that we won't enter a dark forest and there's…

Yep, im sure lots of people have written a lot of random stuff on a lot of forums that should absolutely stay anonymous from gossip to family secrets to honesty about career/workplace and what not. If stylometric analysis runs on all comments on the internet then yeah. Bad things will happen, very very bad. I honestly think it should be at least illegal to do this kind of analysis because it'll be a treasure trove fo…

Idk the state of the art for forensic stylometrics or the precedents but I would be surprised if this hasn’t already been presented as evidence in many (most?) notable jurisdictions. Not definitive evidence but supporting evidence. So far from being declared illegal, it’s probably mostly established as a tool for law enforcement.

Meanwhile gangsters or nation states are probably already working on automated deanonymization and blackmail at scale, but will target the dark web before the public one. Not sure any of this stuff is that changed by the advent of deep fakes and llms tho, probably just boring classical statistics does ok?

Re: They stole my voice with AI

#192

Post by roenxi » Sun, Sep 22, 2024, 10:01 AM UTC

ryzvonusef wrote: post #176

kmlx wrote: post #162

Earlier quoted context omitted.

> what if AI was used to imitate a person saying something blasphemeous? > My country is already has blasphemy lynching mobs in your case the problem is not AI, it’s your country.

Your country might not have lynching mobs, but you can't deny there are certain taboo topics in your society also, certain slurs and other opinions which would take you ages to clense and even then never fully. If an AI fake-porn of some ordinary person involving a minor was unleashed, think of the utter shame and horror they would be treated by people for the rest of their lives, even if it were proven false. No one…

But mob formation and mass social shaming on little evidence already happens without AI and it isn't clear why AI would make it worse. If it gets so easy to create fake videos that'll just muddy the waters and people would trust rumour less, not form mobs more.

It happens once, then some wag creates videos of all the mob leaders blaspheming or whatever. Undermines the idea that the videos are the root cause.

Re: They stole my voice with AI

#193

Post by pjc50 » Sun, Sep 22, 2024, 10:04 AM UTC

kmlx wrote: post #162

ryzvonusef wrote: post #135

Everyone has their own fears about AI, but my fears are especially chilling; what if AI was used to imitate a person saying something blasphemeous? My country is already has blasphemy lynching mobs based on the slightest perceived insult, real or imagined. They will mob you, lynch you, burn your corpse, then distribute sweets while you family hide and issue video messages denouncing you and forgiving the mob. And thi…

> what if AI was used to imitate a person saying something blasphemeous? > My country is already has blasphemy lynching mobs in your case the problem is not AI, it’s your country.

The US equivalent is much less labour intensive than a lynch mob: it's mass shooters radicalized by things they've read on the internet.

Or https://www.npr.org/2024/09/19/nx-s1-5114047/springfield-ohi... , where repeating racial libel causes a public safety problem.

While this kind of incitement in no way requires AI, it's certainly something that's easier to do when you can fake evidence. See also https://www.bbc.co.uk/news/articles/c5y87l6rx5wo

Re: They stole my voice with AI

#194

Post by pjc50 » Sun, Sep 22, 2024, 10:06 AM UTC

latexr wrote: post #167

kmlx wrote: post #162

Earlier quoted context omitted.

> what if AI was used to imitate a person saying something blasphemeous? > My country is already has blasphemy lynching mobs in your case the problem is not AI, it’s your country.

The comment didn’t say the problem was AI, it said they feared its consequences, which is a perfectly valid concern. It’s like if someone said “I’m scared of someone bringing a semi-automatic weapon to my school and doing a mass shooting. My country has lax laws about guns and their proper use”. And then you said “in your case the problem is not guns, it’s your country”. I mean, it’s technically true, but also unhelp…

> It’s like if someone said “I’m scared of someone bringing a semi-automatic weapon to my school and doing a mass shooting. My country has lax laws about guns and their proper use”. And then you said “in your case the problem is not guns, it’s your country”.

The US has both problems: widespread availability of weapons and a high level of freedom to incite violence through spreading lies about groups. Which is why it sees much more of these incidents than countries which have similar levels of gun ownership.

The non-gun version of the problem is mass stabbings, which are less lethal.

Re: They stole my voice with AI

#195

Post by latexr » Sun, Sep 22, 2024, 10:07 AM UTC

blueflow wrote: post #188

latexr wrote: post #174

Earlier quoted context omitted.

That sounds like a dangerous proposition. Either they fabricate allegations about a “nobody” and put them in trouble or they fabricate allegations about those in power and will be investigated and put themselves in danger. Neither strategy is good.

It worked pretty fine with #MeToo. Not saying there was no collateral damage, but the end result was that people now default to not believing such allegations.

I don’t understand what you’re arguing, could you be more specific? I know about #MeToo, but what exactly are you saying worked, and what are you claiming people default to not believing?

Re: They stole my voice with AI

#196

Post by ryzvonusef » Sun, Sep 22, 2024, 10:19 AM UTC

roenxi wrote: post #192

ryzvonusef wrote: post #176

Earlier quoted context omitted.

Your country might not have lynching mobs, but you can't deny there are certain taboo topics in your society also, certain slurs and other opinions which would take you ages to clense and even then never fully. If an AI fake-porn of some ordinary person involving a minor was unleashed, think of the utter shame and horror they would be treated by people for the rest of their lives, even if it were proven false. No one…

But mob formation and mass social shaming on little evidence already happens without AI and it isn't clear why AI would make it worse. If it gets so easy to create fake videos that'll just muddy the waters and people would trust rumour less , not form mobs more. It happens once, then some wag creates videos of all the mob leaders blaspheming or whatever. Undermines the idea that the videos are the root cause.

You are giving too much credence to the analytical capabilites to a lynch mob, and you are underestimating how much AI lowers the barrier for creating realistic 'proof' for incitement.

Also there are no mob 'leaders'. These are everyday folks, these are you and me and my neighbours.

A student pissed at a teacher could create a fake AI video of his teacher at night, and wake up to no pesky teacher and feel no remorse over what happened while he slept.

The student was not a mob leader, just some dumbass who was angry at their teacher for giving bad grades, and had access to simple AI tools (soon available online for a trivial fee, if not free).

____

Be honest, are you 100% sure of the status of that facebook account you deleted in 2018? Zuck never deleted it, that content is still there to be mined and abused. Also not that hard to just log in to the damn thing and post something spicy.

I don't remember if old posts can be edited in FB, cause that would allow the person to really gild the lily.

Re: They stole my voice with AI

#197

Post by benterix » Sun, Sep 22, 2024, 10:22 AM UTC

ryzvonusef wrote: post #135

Everyone has their own fears about AI, but my fears are especially chilling; what if AI was used to imitate a person saying something blasphemeous? My country is already has blasphemy lynching mobs based on the slightest perceived insult, real or imagined. They will mob you, lynch you, burn your corpse, then distribute sweets while you family hide and issue video messages denouncing you and forgiving the mob. And thi…

I'm very sorry to say this but if you live in a country that is killing others for what they say, AI is probably not your biggest problem. And I don't believe an easy solution exists.

Re: They stole my voice with AI

#198

Post by benterix » Sun, Sep 22, 2024, 10:27 AM UTC

roenxi wrote: post #192

ryzvonusef wrote: post #176

Earlier quoted context omitted.

Your country might not have lynching mobs, but you can't deny there are certain taboo topics in your society also, certain slurs and other opinions which would take you ages to clense and even then never fully. If an AI fake-porn of some ordinary person involving a minor was unleashed, think of the utter shame and horror they would be treated by people for the rest of their lives, even if it were proven false. No one…

But mob formation and mass social shaming on little evidence already happens without AI and it isn't clear why AI would make it worse. If it gets so easy to create fake videos that'll just muddy the waters and people would trust rumour less , not form mobs more. It happens once, then some wag creates videos of all the mob leaders blaspheming or whatever. Undermines the idea that the videos are the root cause.

> people would trust rumour less, not form mobs more

I believe you are right. When the first unbelievable clips appeared on Facebook (where the most gullible people are), you could see the old ladies cheering "oh how wonderful! where did you film that?" Now in the comments section you will mostly see people saying "sheesh, stop polluting my feed with these fake AI-generated spam". Don't get me wrong, many people still fall for them, but the tendency seems clear - after all, we managed to survive because of learning.

Re: They stole my voice with AI

#199

Post by fragmede » Sun, Sep 22, 2024, 10:28 AM UTC

pjc50 wrote: post #193

kmlx wrote: post #162

Earlier quoted context omitted.

> what if AI was used to imitate a person saying something blasphemeous? > My country is already has blasphemy lynching mobs in your case the problem is not AI, it’s your country.

The US equivalent is much less labour intensive than a lynch mob: it's mass shooters radicalized by things they've read on the internet. Or https://www.npr.org/2024/09/19/nx-s1-5114047/springfield-ohi... , where repeating racial libel causes a public safety problem. While this kind of incitement in no way requires AI, it's certainly something that's easier to do when you can fake evidence. See also https://www.bbc.co…

https://apnews.com/article/ai-maryland-principal-voice-recor...

In the US, freedom of speech means the government won't stifle you, but the court of public opinion will crucify you for saying the wrong thing. In the linked case, a school principal was framed for being a racist, a fire able offense. Better than being killed by religious extremists, I suppose, but still not great. Thankfully in this case they were able to find the perpetrator who faked evidence, but we have a problem.

Re: They stole my voice with AI

#200

Post by mrkramer » Sun, Sep 22, 2024, 10:28 AM UTC

ryzvonusef wrote: post #135

Everyone has their own fears about AI, but my fears are especially chilling; what if AI was used to imitate a person saying something blasphemeous? My country is already has blasphemy lynching mobs based on the slightest perceived insult, real or imagined. They will mob you, lynch you, burn your corpse, then distribute sweets while you family hide and issue video messages denouncing you and forgiving the mob. And thi…

The only logical legal solution is that any content of you shared by you is legitimate one and all other content of you shared by somebody else is presumed non-authenthic and possibly fake.