On his X account this weekend, Elon Musk shared a video of Vice President Kamala Harris. Or rather, it purported to be of her: taking footage from her new campaign ad, the video swaps out Harris’ voice for a synthetic one that calls President Biden senile, Harris herself “a diversity hire,” and generally seeks to undermine her nascent presidential campaign. Watching the roughly two-minute clip, it seems fairly self-evident that the video is a parody. Indeed, the first account to upload it to X identified it as such — perhaps so as not to run afoul of the platform’s stated policy, which bans deceptive synthetic media. Musk, however, removed the label. Here’s Ken Bensinger at the New York Times: Mr. Musk’s post, which has since been viewed 98 million times, would seem to run afoul of X’s policies, which prohibit sharing “synthetic, manipulated or out-of-context media that may deceive or confuse people and lead to harm.” [...]
The Federal Election Campaign Act prohibits fraudulent misrepresentation of federal candidates or political parties, but the law, written in 1971, is ambiguous when it comes to modern technologies such as artificial intelligence.
In an earlier era of social networking, Musk’s post would have been shocking. In 2019, Facebook and Twitter faced strong criticism and Congressional pressure to act after videos that purported to show then-House Speaker Nancy Pelosi slurring her words began to go viral on the platforms. Now the owner of one of those networks is posting the deepfakes himself. It’s tempting to dismiss the Harris post as more bad behavior from one of the tech world’s most irresponsible actors. (A few hours after posting the deepfake, Musk had already moved on to promoting vaccine skeptics, while making a “deez nuts” joke to California Gov. Gavin Newsom.) But sharing weaponized misinformation in the form of lazy jokes has quickly come to define the developing presidential campaign between Harris and Donald Trump. Across social networks, Democrats and Republicans are flooding the feed with obviously untrue statements about one another and calling it a joke. Welcome to the shitpost election. To shitpost, of course, is to attempt to derail online discussions by posting fake, outrageous, and often low-quality material. And it is not new to this election. Oculus co-founder Palmer Luckey funded a group dedicated to creating shitposts about Hillary Clinton during the 2016 election. In 2019, the leader of the Liberal Democrats in the United Kingdom had to deny that she killed squirrels for fun after a shitpost on the subject went viral. And if the past few days have made anything clear, it’s that political shitposting is not exclusively a province on the right. Anyone who has opened X over the past few days has likely seen dozens of posts “joking” that vice presidential candidate J.D. Vance had sex with a couch. Here’s Olivia Craighead in The Cut: On July 15, Vance was announced as Donald Trump’s running mate. Shortly after, X user @rickrudescalves (whose account is now private) wrote, “can’t say for sure but he might be the first vp pick to have admitted in a ny times bestseller to fucking an inside-out latex glove shoved between two couch cushions (vance, hillbilly elegy, pp. 179-181).” It’s an indictment on Vance’s demeanor that so many people believed this without question, but it is, in fact, a lie.
Pages 179–181 of Hillbilly Elegy are actually about Vance’s early days at Ohio State University. Vance wrote that he remembered every magical detail about arriving on campus for orientation, the first of which is “lunch at Chipotle.” Hell yeah, dude.
But this is a case where the truth takes a (sorry) backseat to the larger political project of painting Vance as a weirdo. And no artificial intelligence or other cutting-edge technology was necessary to vault that piece of misinformation to the forefront of this year’s campaign. John Oliver and Stephen Colbert are already telling couch jokes on their national television shows. And its popularity online is not limited to the anything-goes environment of X: couch memes ran rampant over the weekend among the leftists on Bluesky and even materialized on the famously joke-averse Threads. (Where a prominent art critic became perhaps the 15th person I saw make a comment about “sectional healing.” Of course, there is a long tradition of Americans spreading nasty lies about their candidates for national office. In the 2000s alone we had falsehoods accusing John McCain of having an illegitimate child; John Kerry of having exaggerated his bravery during the Vietnam War; and Barack Obama having been born outside the United States. What those earlier lies shared is that they were intended to be believed. They were promulgated through whisper networks by actors who sought to conceal their role in spreading them. They were dirty tricks. With the Harris clip and the Vance couch memes, though, we’re seeing something different. These aren’t dirty tricks, they’re “jokes.” They’re self-evidently false, although no one who is sharing them would mind if people believed them. They’re bullshit, in the Harry Frankfurt sense: they are intended to persuade without any regard for what the truth might be. “Bullshit” is also a critique that gets leveled at generative artificial intelligence. Tools like ChatGPT similarly have no real relationship to the truth; they confidently answer our questions despite the fact that their underlying models are only making statistical predictions of which token might follow the one previous. Every day now on social media, these threads come together. AI tools make bullshit; bullshit gets turned into shitposts; shitposts take over the national conversation; 2024 becomes the shitpost election. There was a time when tech CEOs were hauled before Congress to answer for this sort of thing: why there was so much misinformation on their networks; how their algorithms promoted its spread; how those algorithms (or the employees making them) might have one political bias or another. That was a logical response to the 2016 election, when Russians were found to have surreptitiously flooded Facebook, Twitter, and YouTube with fake accounts and posts intended to sow division. It was also a logical response to the 2020 election, when President Trump used Facebook and Twitter repeatedly to spread lies about mail-in voting and, eventually, the results of the election. But Americans have gradually become inured to the idea that political actors will attempt to manipulate them on social media and elsewhere. The naivete of 2016 has been replaced with a kind of wry cynicism. As Max Read wrote in 2022: “The thing about misinformation is, it won.” When you have resigned yourself to the ubiquity of misinformation — and when platforms have aggressively downsized the content moderation teams that might otherwise prevent you from doing so — you shitpost. This dynamic could change, over course. More pernicious, less obviously phony synthetic media of political candidates will almost certainly emerge. And the campaign itself may grow more serious as Election Day draws closer, and people begin to post more earnestly about the stakes of winning and losing. Then again, there are few things in politics more serious than an attempted assassination, and Americans just shitposted their way through that, too. Someday, perhaps, platforms will once again face pressure to promote a more truthful version of the world around them. But in the meantime, Americans are voting with the repost button, and for the moment they seem perfectly content with fiction. Governing- Kamala Harris is garnering endorsements from tech industry power players like Sheryl Sandberg, Reed Hastings, and Melinda French Gates. (Clare Duffy / CNN)
- Zoom is becoming an important tool for virtual rallies, especially for Harris’s campaign. (Jacob Knutson / Axios)
- Liberal creators are energized over Harris’s campaign and what it means for memes and digital strategy. (Kevin Roose / New York Times)
- Silicon Valley elites are openly clashing over politics in a way that has rarely been seen in public before. It all started with people getting really tired of David Sacks. (Ryan Mac, Erin Griffith, and Mike Isaac / New York Times)
- “Never sell your bitcoin,” Donald Trump said at the Bitcoin Conference, adding that as president he would ensure the government does not sell its bitcoin holdings. He was previously “not a fan” of crypto while he was in the White House. (MacKenzie Sigalos / CNBC)
- Iran is running covert online influence operations to harm Trump’s campaign, US intelligence officials say. (Dustin Volz / Wall Street Journal)
- TikTok is deploying all the tools it has to fight a US ban, sending lawyers and lobbyists to Washington and spending millions in ad campaigns. (Emily Birnbaum and Sabrina Willmer / Bloomberg)
- Neo-Nazis and white supremacists are trying to use TikTok to recruit new members by spreading extremist propaganda, a new report found. (David Gilbert / WIRED)
- ByteDance had access to a search tool inside Lark, its internal communications program, that could have collected information on US TikTok users, the DOJ alleges. (Victoria Bisset and Drew Harwell / Washington Post)
- US border agents must obtain a warrant before searching the electronic devices of travelers crossing the border, a federal judge ruled. (Zack Whittaker / TechCrunch)
- Anthropic is aggressively scraping data from websites to train its AI systems, Freelancer.com and other web publishers say. Ugh. (George Hammond / Financial Times)
- X is training Grok on user data by default, and the option can only be disabled from the web app. (Kevin Okemwa / Windows Central)
- Open source AI models are more likely to promote competition, as they allow developers to customize with fewer restrictions, FTC chair Lina Khan said. Somewhat surprising to hear from Khan! (Leah Nylen and Lizette Chapman / Bloomberg)
- Apple agreed to a set of voluntary AI safeguards by the Biden administration, joining other big tech companies. (Akayla Gardner / Bloomberg)
- Apple reached a tentative union deal with retail employees in Maryland, which will increase pay by an average of 10 percent. (Josh Eidelson / Bloomberg)
- Google fixed a security vulnerability that allowed criminals to bypass email verification to create Google Workspace accounts and impersonate a domain holder at third-party services, the company said. (KrebsonSecurity)
- Talkie, number five among the most-downloaded free entertainment apps in the US, is reportedly owned by a Chinese startup. (Raffaele Huang / Wall Street Journal)
- AI technology in China is quickly closing the gap on the US with open-source code and quick distribution to consumers. (Meaghan Tobin and Cade Metz / New York Times)
Industry- Elon Musk’s transgender daughter, Vivian Jenna Wilson, said that when she was a child Musk berated her for being queer and feminine, (David Ingram / NBC News)
- A look at Linda Yaccarino’s rise, her management of X, and her Elon Musk problem. The company's revenues continue to be in free fall. (Kate Conger / New York Times)
- xAI’s goal to build one of the world’s largest supercomputers in South Memphis is drawing mixed reactions from locals. (Patrick Sisson / Bloomberg)
- CapCut, ByteDance’s video-editing app, is increasingly poaching users away from Adobe and Canva, despite a looming ban. (Brody Ford and Alicia Clanton / Bloomberg)
- Meta is rolling out AI Studio, a tool for Instagram creators to design personas that can answer questions and chat with other users on their behalf. (Karissa Bell / Engadget)
- Apple’s AI features will miss the initial iPhone and iPad software update rollout, and will reportedly be part of October’s updates instead. (Mark Gurman / Bloomberg)
- Amazon is still reportedly losing money on its biggest acquisition – gamer streaming company Twitch. (Salvador Rodriguez, Sarah E. Needleman and Sebastian Herrera / Wall Street Journal)
- Twitch’s new mobile app update prioritizes a TikTok-like video feed, complete with an algorithmic recommendation feed instead of a “following” feed. (Jay Peters / The Verge)
- Google’s Gemini AI ad about encouraging a kid to use AI to craft a fan letter to an Olympian is deeply misguided. Another AI messaging flop from Google. (Anthony Ha / TechCrunch)
- Platforms like Pinterest, LinkedIn, and TikTok are teaming up with news publishers to sell ads. (Sahil Patel / The Information)
Those good postsFor more good posts every day, follow Casey’s Instagram stories. (Link) (Link) Talk to usSend us tips, comments, questions, and shitposts: casey@platformer.news and zoe@platformer.news.
|