Outliner Software Forum RSS Feed Forum Posts Feed

Subscribe by Email

CRIMP Defined

 

Tip Jar

OT - but maybe not entirely - warning about AI assisted deep fakes

< Next Topic | Back to topic list | Previous Topic >

Posted by Stephen Zeoli
Jul 17, 2023 at 02:41 PM

 

Hi, all,

This article warns about the dangers of AI created deepfake photos and videos, and provides some advice about how to guard against them. I imagine most if not all of you are tech savvy to the point that you are already sensitive to these things, but I think it is important to get this information out to everyone.

https://theconversation.com/events-that-never-happened-could-influence-the-2024-presidential-election-a-cybersecurity-researcher-explains-situation-deepfakes-206034

Steve

 


Posted by Amontillado
Jul 17, 2023 at 03:34 PM

 

Stephen Zeoli wrote:
Hi, all,
> >This article warns about the dangers of AI created deepfake photos and
>videos, and provides some advice about how to guard against them. I
>imagine most if not all of you are tech savvy to the point that you are
>already sensitive to these things, but I think it is important to get
>this information out to everyone.

It’s scary. On the one hand, it’s just more polished trick photography, but on the other hand the verisimilitude can be dangerously compelling.

And on the third hand, since this is AI, the application of simulation in channels of limited resolution is scary. For instance, the mother compelled to send money to a kidnapper after hearing an AI simulation of her daughter’s voice.

The world will come full circle when AI is used for authorization and humans learn to simulate AI. I can just imagine the angst. “You idiot! You just wire transferred $10 million to a hacker!”

“But boss, I had AI authorization!”

“No, you moron! That was Rich Little impersonating Richard Nixon, neither of which is even president any more!”

 


Posted by Andy Brice
Jul 20, 2023 at 11:03 AM

 

Haven’t read the article yet, but aking has been with us probably as long as human history. Even before AI deepfakes you could edit a video for your own ends (e.g. take a quote out of context). So I think its more a matter of degree than a new thing. It is a bit scary how easy it is to create these deepfakes though.

It will be interesting to see how well forensic tools can keep up with the new technology

 


Back to topic list