Don’t Let Them Steal Your Election

🌈 Abstract

The article discusses the potential dangers of AI technology, particularly in the context of the upcoming US elections. It highlights how AI-powered tools like deepfakes, bots, and algorithmic feeds can be weaponized to spread misinformation and influence the democratic process. The article also raises concerns about the role of social media platforms, such as Twitter and TikTok, in amplifying these threats.

🙋 Q&A

[01] Deepfakes and Subtle Propaganda

1. What are deepfakes, and how can they be used as propaganda vectors?

• Deepfakes are fake images or videos created using deep learning techniques, which can range from obvious collages to highly realistic and subtle manipulations.
• The article states that the most dangerous deepfakes are the subtle ones that are hard to detect, as they can be used to twist the meaning of a speaker's words and enrage a large number of people.
• Even obvious deepfakes can be powerful as a means of expression, revealing truths about the state of technology and society.

2. How can people be fooled by the latest AI-generated images?

• The article argues that most people would be fooled by the latest AI-generated images, as the technology has advanced to the point where it can create highly realistic and convincing visuals.
• The article cautions against the "survivorship bias" of thinking one can always tell a fake, as the capabilities of AI continue to improve rapidly.

[02] Bots and the Information Ecosystem

1. How can bots be used to influence elections?

• The article states that bots can be used to flood social media with propaganda, and that AI works well as fuel for national ideological propaganda campaigns.
• It notes that the cost of using AI models has dropped to zero, while the quality remains high, making them a powerful and inexpensive tool for bad actors to deploy.

2. What is the impact of bots on the information ecosystem?

• The article suggests that once bots spread and pollute the information ecosystem, the "trust first, check later" phase of the web will be over.
• It cites statistics showing that bots already account for a significant portion of internet traffic, with a significant percentage being "bad bots" used for scamming, spamming, and other malicious activities.

[03] Social Media Platforms and Algorithmic Influence

1. How can social media platforms like TikTok and Twitter be used to influence elections?

• The article raises concerns about TikTok's owner company, ByteDance, being Chinese, and the potential for the platform to be used to expand China's global influence operations and undermine U.S. democracy.
• Regarding Twitter (now X), the article criticizes Elon Musk's control over the platform's algorithm and his public endorsement of Donald Trump, which it argues could amount to direct or indirect interference in the elections.

2. What is the role of algorithmic feeds in amplifying misinformation and political propaganda?

• The article argues that the algorithmic feeds used by social media platforms are designed to keep users engaged, often by promoting "rage bait" content that can distort people's perceptions of the world around them.
• It suggests that these algorithmic feeds are the "worst invention of the 21st century" in terms of their potential to influence political discourse and the democratic process.

[04] The "Liar's Dividend"

1. What is the "liar's dividend," and how can it be used to spread disinformation?

• The "liar's dividend" is the opposite of a deepfake, where someone calls out true information as false, ascribing blame to AI technology.
• The article provides the example of Donald Trump accusing Kamala Harris of using AI to create a fake crowd, even though the crowd was real, as a way to plant doubts among his followers using a somewhat plausible lie.
• The article suggests that the "liar's dividend" is a hidden ace up the sleeve of the deceitful and untrustworthy, as it allows them to sow confusion and distrust.