This article is from a special report on the Athens Democracy Forum, held in association with The New York Times, where experts gathered in the Greek capital last week to discuss global issues.
Lorena Austin, a U.S. Democrat elected to Arizona’s House of Representatives in 2022, fights for issues such as housing and homelessness, education and climate.
Yet for a brief moment in September, voters were left with a completely different impression of her.
In a social-media campaign ad (posted by the conservative advocacy group Turning Point Action), a smiling Ms. Austin was pictured alongside a group of Democrats supporting the other side: the conservative Republican candidate Andy Biggs, who is running for governor of Arizona in November.
The image was pure fabrication.
Technology and its threats to democracy were the central theme of last week’s Athens Democracy Forum, an annual gathering of policymakers, business leaders, academics and activists held in association with The New York Times.
The case of Ms. Austin’s fake photo came up in a presentation by William Gates, executive director of the Mechanics of Democracy Laboratory at Arizona State University, who gave it as an example of the mounting dangers to democracy from disinformation created by artificial intelligence.
He was hardly alone. At the Athens gathering, one panel after another warned of the potential risks that social media, deepfakes, cyberattacks and the rapid advances of A.I. posed to society, democracy and humanity as a whole.
Other key topics at this year’s forum: the United States’ ongoing war with Iran, and the aftermath of the Nov. 3 U.S. congressional elections.
In an interview after his presentation, Mr. Gates acknowledged that lying about your opponent has been going on for as long as we’ve had politics.
But what has changed, he said, was how fast troublesome fakes could be created. He pointed out that in the past, creating a fake photograph was a time-consuming job requiring a graphic designer and a good deal of expertise. Not today, he said. All it took today was asking A.I. to do it for you.
But the damage can be extensive. In a fake video, for example, an election official might declare that an election has been canceled, urging voters to stay home, Mr. Gates said; or announce that 50 polling stations have closed, sending voters to nonexistent locations.
As in previous years, the forum gave voice to the young — university students as well as teenagers — to ensure that the discussions were of relevance to their generation.
One youth-related panel focused on teen access to social media, following Australia’s decision last December to ban children under 16 from having accounts on major social media platforms.
Speaking in favor of controls was Peter G. Kirchschläger, a professor of ethics at the University of Lucerne in Switzerland and a visiting professor at ETH Zurich. He said young people, now “addicted” to social media, needed “rehab measures.”
“We would never distribute cigarettes to 10-year-olds and say, ‘Look: find a responsible way to deal with tobacco,’” he said.
He called for “preventive measures” and social-media education to ensure that young people would “use these technologies in a responsible way.”
But Alexia Kadoglou, a teenage student, said she was against social media bans, because she used social media every day to communicate with her friends, keep up with the news and express herself.
“We should have access to social media without any restrictions, but we should learn how to be digitally responsible,” she said. “Young people should have a say.”
Besides their harmful impact on adolescents, the internet and social media have also done demonstrable damage to journalism: Readers have migrated online, and revenues from advertising and print subscriptions have dwindled. Now, A.I. is throwing up a whole new set of challenges, said speakers in a panel on rethinking journalism.
“It’s an enormous threat to our industry,” said Martin Baron, a former executive editor at The Washington Post. He said many people got their news from chatbots, which they trusted more than news organizations because they assumed the news was gathered from all sources and was not biased in any way. And yet, he said, “A.I. can often produce really inaccurate results.”
Eliza Anyangwe, former editor in chief of Fuller, a global nonprofit news organization reporting on women and gender-diverse people, said that when reporting on A.I., journalists often gave ample voice to the heads of the big tech companies and ended up “amplifying the same tropes” and messages that the technology bosses were putting out.
“The challenge of dealing with A.I. is: How do we maintain some distance of it, where we are critical of it, where we take a breath to think and where we are engaging audiences” and asking if they understand what A.I. is, she said.
The powers and perils of A.I. were spelled out in detail by The Times’s columnist Thomas Friedman, who is writing a book about the technology.
In an opening-night address, he described it as “the second Big Bang” that had been “set off by us: We have become godlike as a species.”
He said humankind had, through the advances of science and technology, created an artificial brain, changed the climate, generated an entirely new interconnected galaxy called cyberspace, produced weapons that could destroy the planet, and allowed the convergence of quantum computing, fusion energy and A.I.
“We now have the power to part the Red Sea. Just one problem: without the Ten Commandments,” said Mr. Friedman in an address held inside the neoclassical headquarters of the Academy of Athens.
“The first challenge for democracies is going to be to collaborate globally” and “collectively write these together,” he said. It was, Mr. Friedman added, a do-or-die necessity: “We are going to rise together, or we’re going to fall together.”
That prognosis was echoed by Finance Minister Kyriakos Pierrakakis of Greece, who in a subsequent conversation urged the world’s governments to get together to regulate A.I.
No country in the world is big enough to govern A.I. alone, not even the U.S. or China, said Mr. Pierrakakis, a graduate of Harvard University and of the Massachusetts Institute of Technology. “We need global governance.” Governments are “not yet fully up to speed with the change that is happening,” and with the velocity of that change. What previously took years or decades now happens in months or even weeks, said Mr. Pierrakakis, who previously served as Greece’s Minister for Digital Governance.
He said institutions are “the software of our democracy” and concluded that “the software needs to change to adapt to the A.I. era.”
As ever at the forum, geopolitics and international conflicts were in the spotlight. This year’s focus was Iran — which the United States and Israel attacked on Feb. 28, decapitating its leadership. Since then, Iran has led missile and drone attacks on countries and ships in the region, pushing up oil prices.
Speaking on the Iran panel, Mr. Friedman from The Times predicted that the conflict would end in a U.S.-Iran negotiation, but “given the weak position the Americans have now, the negotiation will end in Iran’s favor. We cannot end this war, we the United States, without giving a new lease on life to the Islamic Republic in the near term.”
Nazenin Ansari — managing editor of the digital publications Kayhan London and Kayhan Life, read by the global Iranian community — said: “The survival of the Islamic Republic and the well-being of the Iranian people are profoundly different things.”
She said that while international attention is focused on the Islamic Republic, the Iranian people, “whose voices are so often missed,” are enduring economic hardship and severe repression.
The post Debating the Risks Posed by Artificial Intelligence appeared first on New York Times.
