Artificial intelligence helped predict voter turnout in Mississippi last year when a group used the technology to transcribe, summarize and synthesize audio recordings of their door knockers’ interactions with voters to create reports of what they heard in each precinct.
Another group recently compared human- and AI-translated messages in six Asian languages and found they were all similarly effective. A Democratic company tested four versions of a voice-over ad—two spoken by humans, two by AI—and found that the male AI voice was just as convincing as its human equivalent (the female voice outperformed its AI equivalent).
The age of artificial intelligence has officially arrived in the election campaign. But the much-anticipated and feared technology remains confined to the fringes of American campaigns.
With less than six months until the 2024 election, the political use of AI is more theoretical than transformative, both as a constructive communication tool and as a means of spreading dangerous disinformation. The Biden campaign said it has strictly limited its use of generative AI – which uses prompts to create text, audio or images – to productivity and data analysis tools, while the Trump campaign said it does not use the technology at all.
“This is the dog that didn’t bark,” said Dmitri Mehlhorn, political adviser to one of the Democratic Party’s most generous donors, Reid Hoffman. “We haven’t found a cool thing that uses generative AI to invest in and actually win elections this year.”
Mr. Hoffman is hardly an AI skeptic. He previously served on the board of Open AI and recently conducted an “interview” with an AI version of himself. For now, however, the only political applications of the technology that deserve Mr. Hoffman’s money and attention are what Mr. Mehlhorn calls “unsexy productivity tools “ referred to.
Eric Wilson, a Republican digital strategist who runs a campaign technology investment fund, agreed. “AI is changing the way campaigns are run, but in the most boring and mundane way imaginable,” he said.
Technologists and political activists have little doubt about the power of AI to change the political scene – eventually. A new report from Higher Ground Labs, which invests in political technology companies to advance progressive causes and candidates, found that while the technology is still in the “experimental stage,” it also represents “a generational opportunity” for the Democratic Party represent moving forward.
So far, the Democratic National Committee has conducted more modest experiments, such as using AI to detect anomalous patterns in voter registration records and finding notable voter removals or additions.
Jeanine Abrams McLean, the president of Fair Count, the nonprofit that led the AI experiment in Mississippi, said the pilot included 120 voice notes recorded after meetings with voters and transcribed by AI. The team then used the AI tool Claude to map geographic differences in opinion based on what customer surveyors said about their interactions.
“Synthesizing the voice notes with this AI model showed us that sentiment from Coahoma County was much more active, suggesting a plan to vote,” she said. “Whereas in Hattiesburg we didn’t hear the same sentiments.”
In fact, she said, voter turnout was down in the Hattiesburg area.
Larry Huynh, who oversaw the AI voice-over advertising, said he was surprised at how well the AI voices had performed. He and most of his colleagues at Democratic consulting firm Trilogy Interactive had thought the male AI voice sounded “the most stilted.” Nevertheless, according to tests, it proved to be convincing.
“You don’t necessarily have to have a human voice to make an effective ad,” said Mr. Huynh, who, as current president of the American Association of Political Consultants, thinks a lot about the ethics and economics of AI technology. Still, he added, tinkering with models to create a new AI voice was just as labor-intensive and costly as hiring a voice actor.
“I don’t think so,” he said, “it actually saved us money.”
Both Democrats and Republicans are also vying to protect themselves against the threat of a new category of political dark arts that includes AI-powered disinformation in the form of deepfakes and other false or misleading content. Ahead of the New Hampshire primary in January, an AI-generated robocall that mimicked President Biden’s voice to suppress votes led to a new federal regulation banning such calls.
For regulators, lawmakers and election officials, the incident highlighted their disadvantages in dealing with even inexperienced troublemakers who can move more quickly and anonymously. The fake Biden robocall was made by a magician in New Orleans who holds world records for fork bending and escaping from a straitjacket. He said he used an off-the-shelf AI product that took 20 minutes and cost a dollar.
“What was concerning was the ease with which a random member of the public who really didn’t have a lot of experience with AI and technology could initiate the call themselves,” New Hampshire Secretary of State David Scanlan told a Senate committee hearing on the role of AI in the election this spring.
AI is “like a match to gasoline,” said Rashad Robinson, who helped write the Aspen Institute’s report on information disruption after the 2020 race.
Mr. Robinson, the president of Color of Change, a racial justice group, described the kind of “nightmare” scenario that he said would be all but impossible to stop. “You can have the voice of a local pastor calling three thousand people and telling them, ‘Don’t go to the polls because there are armed white men there and I’m fighting for an extra voting day,'” he said. “The people who develop the tools and platforms that make this possible have no real responsibility and no real consequences.”
The prospect of similar 11th-hour AI-powered disruptions is causing New Mexico Secretary of State Maggie Toulouse Oliver to lose sleep. Ahead of her state’s primary, she launched an ad campaign warning voters that “AI won’t be so obvious this election season” and advises, “When in doubt, give it a try.”
“We often find ourselves behind the eight ball during elections,” she said, adding: “And now we have to deal with this new wave of activity.”
AI has already been used to mislead campaigns abroad. In India, an AI version of Prime Minister Narendra Modi addressed voters by name on WhatsApp. In Taiwan, an AI representation of outgoing President Tsai Ing-wen appeared to encourage cryptocurrency investments. In Pakistan and Indonesia, dead or imprisoned politicians have reappeared as AI avatars to appeal to voters.
So far, most counterfeits have been easily debunked. But Microsoft’s Threat Analysis Center, which studies disinformation, warned in a recent report that deepfake tools are becoming more sophisticated by the day, even though one that could influence American elections “likely has not yet come to market.”
In the 2024 race, many candidates are approaching artificial intelligence cautiously, if at all.
According to a statement from Steven Cheung, a spokesman, “the Trump campaign does not engage in or use AI.” However, he said that the campaign, “like many other campaigns across the country, uses a range of proprietary algorithmic tools to Deliver emails more efficiently and prevent registration lists from being filled with incorrect information.”
But the Trump campaign’s reluctance to embrace AI hasn’t stopped his supporters from using the technology to create fake images of the former president surrounded by Black voters, a constituency he is aggressively courting.
The Biden campaign said it has placed strict limits on the use of AI. “Currently, the only authorized campaign use of generative AI is for productivity tools such as data analysis and industry-standard coding assistants,” said Mia Ehrenberg, a spokeswoman for the campaign.
A senior Biden official, granted anonymity to discuss internal operations, said AI is most commonly used in the campaign to create behind-the-scenes efficiencies, such as testing which marketing messages lead to clicks and other forms of interaction, a process known as conversation marketing. “Not the stuff of science fiction,” the official added.
Artificial intelligence has taken such a central place in the zeitgeist that some campaigns have found that just using the technology helps draw attention to their messages.
After the National Republican Congressional Committee showed AI-generated images of national parks as tent cities for migrants last year, a wave of reporting followed. In response to a recording released by the former president’s daughter-in-law, Lara Trump (her song was called “Anything is Possible”), the Democratic National Committee used AI to create a dissident track about Ms. Trump and made fun of GOP supporters. Sensational, which caught the attention of celebrity gossip site TMZ.
However, digital policy strategists are still figuring out how well AI tools actually work. While many are mundane data processing efforts, some include novel ideas, such as an AI-powered eye contact tool designed to prevent the person in a video from breaking eye contact, which could streamline the recording of scripted videos. With the White House blocking the release of audio recordings of Mr. Biden’s interview with a special counsel, Republicans could instead use an AI-generated recording of Mr. Biden reading the transcript to dramatic effect.
“I don’t know a single person who hasn’t tried to pre-write their content,” said Kenneth Pennington, a Democratic digital strategist, about using generative AI to write early drafts of fundraising messages. “But I also don’t know many people who thought the process was useful.”
In Pennsylvania, a congressional candidate used an AI-powered phone banking service to conduct interactive phone conversations with thousands of voters.
“I share everyone’s deep concern about the potential nefarious use of AI in politics and elsewhere,” candidate Shamaine Daniels said on Facebook. “But we also need to understand and take advantage of the opportunities this technology offers.”
She finished the competition in distant third place.
Source link
2024-05-23 10:39:05
www.nytimes.com