"Personalized 1:1 persuasion, combined with high-quality generated media, is going to be a powerful force."
Once again, OpenAI CEO Sam Altman — whose November launch of the popular chatbot ChatGPT arguably kicked off the widespread public adoption of generative AI technology — is worried about the impacts of the technology that he's actively working to build.
This time, he's concerned about the potentially destructive effects that generative AI tools may have on the democratic process, arguing that high-quality and hyper-targeted synthetic media stand to be a powerful and potentially damaging force in future elections — a fear that's already been raised by a number of experts and industry leaders.
"I am nervous about the impact AI is going to have on future elections (at least until everyone gets used to it)," Altman wrote in an X-formerly-Twitter post on Thursday, "personalized 1:1 persuasion, combined with high-quality generated media, is going to be a powerful force."
It's a fair concern. AI-generated media has already been used in multiple American campaign ads related to the 2024 election, most notably by the likes of the Republican National Committee and the campaign of presidential hopeful Ron Desantis, which used AI tools to generate fake images of an imagined China-US invasion and President Joe Biden smooching former US pandemic response figurehead Anthony Fauci, respectively.
AI-spun misinformation has also already proved itself to be pretty convincing. Though it was quickly debunked as fake news, a likely AI-generated photo of a bombing nearby the Pentagon, for instance, caused a brief but notable dip in the stock market after going viral on Twitter in May.
And back in June, synthetic images of demonic children's clothes at Target sent far-right social media groups spiraling into a culture war tizzy — an incident that seems to speak to Altman's warning about personalized persuasion. People tend to seek out information that confirms what they already believe; with the right prompt and the right platform, it might be relatively easy to suck certain ideological groups into synthetic misinformation traps.
Of course, considering that Altman is profiting off the tools he regularly warns are dangerous, it's sometimes hard to take him seriously when he talks about AI risks, even when he's objectively correct to worry. But as he's said himself in the past, it's probably better that he harbors some fear towards his creations, rather than ignore any risks outright.
"although not a complete solution, raising awareness of it is better than nothing," he wrote in that Thursday thread on X. "we are curious to hear ideas, and will have some events soon to discuss more."
More on Sam Altman: OpenAI CEO Hopeful World Leaders Will Save Us From AI He's Building
Share This Article