It's 2024, and AI is playing a bigger role in our lives than ever before—and nowhere is that more apparent than in politics. This election cycle has seen the rise of AI-generated content, especially deep fakes, which are reshaping the landscape of political discourse. The question isn't just about how these AI tools are being used, but also how well voters can recognize them and understand their potential impact.

Listen to the two chirpy AI hosts of Deep Learning with the Wolf discuss how to spot a deepfake. (Hint: sometimes the secret to detection lies in the skin and the glasses.)

How Much Do Voters Know AI?

Let's face it: not everyone is an AI expert. In fact, many voters may not even realize just how much of what they see and hear online is influenced by machine-generated content. This election, AI-generated videos, audio clips, and even news articles are popping up everywhere. But do voters know what to look for? Do they understand that the video of a candidate saying something controversial might be completely fabricated?

Let's ask Deep fake Jake to show us how easily this technology can be faked and how convincing it is to the average viewer.

The awareness gap is a major concern. A recent study found that only about 40% of voters could correctly identify a deep fake video when shown side by side with genuine footage. And with the AI-generated content becoming increasingly convincing, that number could easily get worse.

Can Voters Recognize AI?

The ability of voters to discern between real and AI-generated content is crucial for a functioning democracy. If voters can't trust what they see, the foundation of informed decision-making starts to crumble. While platforms like Facebook, X, and YouTube are introducing new tools to label AI-generated content, the responsibility still falls heavily on voters to scrutinize what they consume. The trouble is, even those who are aware that deep fakes exist may still be easily misled when under the influence of a powerful narrative or emotionally charged content.

Candidates are also at the mercy of these digital forgeries. A single convincing deep fake, released at the right moment, could sway public opinion or derail a campaign. The threat is so real that many campaigns are now including “AI response teams” specifically tasked with debunking false content.

What Needs to Change?

Education is the first step. Teaching people to spot deep fakes could help mitigate their impact, but it’s not a perfect solution—AI is evolving too quickly for a static approach. A more dynamic strategy might involve a mix of technical solutions (like watermarks or blockchain for verifying authenticity) and legislative oversight to hold creators of malicious deep fakes accountable.

Final Thoughts

AI is here to stay in politics, but our society is still figuring out how to deal with the ethical, social, and practical implications. As we move forward, it’s important for voters to not only be skeptical but also proactive in seeking out credible information.

Additional Resources for Inquisitive Minds:

These two resources were used as the basis for the "Deep Learning with the Wolf" podcast. (a.k.a. the perky discussion by our NotebookLM AI podcasters.):

MIT Media Lab - Detect Fakes - An interactive tool to help individuals learn to identify deep fakes by practicing with real vs. AI-generated videos. media.mit.edu/projects/detect-fakes.

AI in Precision Persuasion. Unveiling Tactics and Risks on Social Media. NATO StratCom Centre of Excellence - Reports and research on deep fakes and their impact on elections and public discourse. (August 20, 2024.)

Vocabulary Key

Deep Fake: A media piece (e.g., video, audio) created using deep learning techniques to mimic real individuals' appearances or voices. Often used to spread misinformation.

AI-Generated Content: Text, images, video, or audio generated by artificial intelligence models, designed to mimic human-created content.

Discourse: The communication or debate surrounding a particular topic, in this case, political discussion during an election.

Narrative: The underlying message or storyline presented to an audience, often used in political content to shape public perception.

Watermark: A mark embedded into media to prove authenticity and prevent forgery. Some have proposed using digital watermarks to label AI-generated content.

Blockchain: A secure digital ledger that could be used to verify the authenticity of media and track whether it’s AI-generated.

AI Response Teams: Campaign groups specifically formed to monitor and respond to AI-generated misinformation, such as deep fakes that could damage a candidate’s reputation.

FAQs

What is a deep fake and why are they important in this election? Deep fakes are AI-generated videos or audio designed to look and sound like real people. They’re important in this election because they can be used to spread misinformation, manipulate public opinion, or damage a candidate’s image.

How can I tell if something is a deep fake? Deep fakes are getting harder to identify, but there are still some signs to watch for: unnatural facial expressions, strange audio syncing, or inconsistencies in lighting. Tools are being developed to help detect them, but ultimately, viewers should approach unexpected or sensational content with skepticism.

Are there laws against deep fakes? Some jurisdictions have started to introduce legislation to hold creators of malicious deep fakes accountable, especially those intended to cause harm. However, regulation is still catching up to the technology, and enforcement varies widely.

How is AI being used by political campaigns? Political campaigns use AI to target ads, craft messages, and sometimes even generate content like speeches or videos. But AI is also being used to defend against misinformation, including deep fakes, by quickly identifying and debunking them.

Why is it important for voters to be aware of AI-generated content? If voters can't tell the difference between real and AI-generated content, they might make decisions based on misinformation. Awareness helps voters verify the credibility of what they see and hear, ensuring they make informed choices.

What are platforms like Facebook and YouTube doing to combat deep fakes? Many platforms are implementing AI-based detection tools to identify and label deep fakes. Some also have policies against spreading manipulated media, although the effectiveness of these measures can vary.

Is there any way to verify that a piece of content is real? A good practice is to cross-check information with multiple reliable sources. There are also emerging technologies like blockchain and digital watermarks designed to prove the authenticity of media.

#deepfakes #ai #election2024 #misinformation #aitechnology #politicaldiscourse #deepfakeawareness #votereducation #digitalmanipulation #aiethics

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit dianawolftorres.substack.com