#夏日生活打卡季 #
The presidential election in the United States has always been a big war full of drama, and all kinds of black materials, hype, rumors, debates, voting, vote counting, lawsuits, and protests have attracted the attention of the world. But if you think you've seen the craziest one, you're dead wrong. Because the 2024 presidential election will face an unprecedented challenge - artificial intelligence deep fraud.
What is deep counterfeiting? Simply put, it is the use of artificial intelligence algorithms to create fake videos and audio that look very real based on a large amount of online video and audio material, so that people can see or hear things they have never said or done.
This technique was initially used to make spoofs and involuntary pornography, such as swapping the faces of female celebrities into. But in recent years, with the development and popularization of technology, deep fakes have also been increasingly used as a political tool to spread disinformation.
For example, on May 23, 2023, a deep-fake video went viral online, disguised as an online endorsement video of Hillary Clinton. This surprised many people because she is a former Democratic presidential candidate, while in the video she is supporting Republican Ron DeSantis, an ally and rival of Trump.
"I actually like Ron DeSantis," Mrs. Clinton said, "and he's the kind of person this country needs, and I mean it." ”
The video looks very real, with Hillary's voice, expressions, and movements exactly like herself. However, if you look closely at the video, you will notice that Hillary's lips and voice do not correspond very well. This is one of the flaws in deep counterfeiting.
The video was actually made using an artificial intelligence tool called DeepFaceLab. It can generate a fake video that can say anything based on anyone's online video and audio material. You just have to type in what you want them to say, choose a suitable template, and wait a few minutes to get a video that looks real.
The tool, launched earlier this year, is one of a new series of "generative AI" tools that make it cheap and easy to make deep fakes. Reuters spoke to a number of experts from areas such as artificial intelligence, online disinformation and political activism.
"It will be very difficult for voters to distinguish between real and false. You can imagine how Trump supporters or Biden supporters use this technology to make opponents look bad. Darrell West, a senior fellow at the Center for Technology Innovation at the Brookings Institution, said.
"There may be something that comes up on the eve of the elections and no one has the opportunity to take them down."
When tools for deepfaking are released, there are not many or perfect precautions to prevent harmful disinformation. Because the tech industry is in an AI race. Aza Ruskin, co-founder of the Center for Human Technology, said.
Former President Donald Trump will compete with DeSantis and others for the Republican nomination and Biden. Earlier this month, he shared a doctored video of CNN anchor Anderson Cooper on his social media platform, Truth Social.
Cooper said in the video: "President Donald J. Trump turned us into a new bastard at CNN's presidential livestream." But if you look closely at the video, you will find that his words do not match the movements of his lips.
CNN called the video a deep fake. Trump's representative did not respond to a request for comment on the video, which remains on his son, Donald Jr., on his Twitter page this week.
According to DeepMedia, a company that works to detect synthetic media tools, the number of various video deepfakes published online has tripled this year and the number of voice deepfakes has increased eightfold compared to the same period in 2022.
DeepMedia estimates that by 2023, about 500,000 video and voice deepfakes will be shared on social media sites worldwide. Until the end of last year, cloned speech cost $10,000 in server and AI training, the company said, but now startups cost just a few dollars.
In order to deal with the threat posed by deep counterfeiting, some organizations and individuals are taking some measures.
For example, some experts have suggested that more regulation of this behavior should require all those who use deep-counterfeiting technology to create or spread false information to add a prominent logo or watermark to the video or audio and remove or correct it within 24 hours. At the same time, the platform needs to implement a moderate mechanism to warn users to pay attention to deep counterfeiting. Strengthen user education to make users aware of the existence and impact of deep counterfeiting.
These measures are commendable and supportive, but will they be effective in preventing or reducing the impact of deep fraud on the 2024 presidential election?
It's hard to say. Because deep counterfeiting technology is constantly evolving and improving, and some people may still use it to mislead or manipulate public opinion for political, economic or other motivations.
Therefore, the most important and effective weapon in this battle with AI technology may be humans themselves - everyone should be vigilant and critical, do not trust or forward any information without reliable sources or evidence, and actively seek and disseminate the truth.
Please click "Follow" in the upper right corner of the page, so that you can see more wonderful content and more high-definition high-quality pictures to learn more useful information. Thank you!