NC ad campaign calls for AI regulations to ‘protect what’s human’

An AI advocacy group has launched an $8 million ad campaign in North Carolina and four other states to persuade lawmakers to act before it is too late.

A national nonprofit launched a million-dollar ad campaign in North Carolina last month to convince state lawmakers they were running out of time to regulate the AI industry and to “protect what’s human” in our relationships, values, and workforce.

The nonprofit, The Future of Life Institute (FLI), is spending $1.2 million in the state on television, streaming and online ads pushing “commonsense safety standards” for a mostly unregulated technology that has already upended nearly every space in which humans interact. The North Carolina campaign is part of an $8 million total effort across four other states, including Iowa, Michigan, Kentucky, and Maine. The ads went live on Feb. 9.

The ad’s message is that without guardrails, AI will very soon get so good at saving humans time and effort, it will make them irrelevant.

It’s a threat that most people relying on the technology to send emails, do their homework, and give them advice don’t fully grasp, and the window to prevent the worst case scenario is far shorter than people realize, Anthony Aguirre, Future of Life’s CEO told Cardinal & Pine before the ad campaign began.

“ AI is specifically being targeted as a human replacement,” Aguirre said, “not just in jobs … but even as companions, as therapists, as lovers, everything.”

And this future is now, he said.

“ The technology is just moving unbelievably quickly, and we’re woefully unprepared for how that’s going to shake everything up and [for] the risks that it will pose to individuals, to families, to society, to the human race as a whole.”

The threat  ”is coming in like a freight train,” he said.

“We’ve got a matter of a year or two to get a handle on things.”

Not a good path for humanity’

It’s hard for an actual freight train to sneak up on whoever it mows down, but that is what is happening with AI, Aguirre said.

The technology has become so ubiquitous so fast, it’s hard for the people who rely on it to see what’s coming. The jobs that AI makes easier today may not exist a year from now. The chatbots that make a teenager feel less lonely on Monday could encourage him to end his life by Friday. There have been cases of deaths by suicide associated with chatbot use. The family of a man who died by suicide filed a lawsuit against Google in March after his heavy use of its chatbot before his death.

This is neither paranoia nor science fiction, and regulation is needed now.

“AI systems are useful tools, some of them really are,” Aguirre said.

 “I enjoy using some of these systems and get utility out of them, but [AI companies] also pushing things like chatbots for kids, and chatbots for everybody, that are built like social media to maximize not just engagement and attention, but even emotional attachment.”

That, he added, “is  not a good path for humanity.”

The ads

The two ads are titled “Wisdom” and “Hands.” They show children riding bikes, playing guitar, and reading. They show farmers running their hands over wheat, carpenters cutting wood, and parents talking to their children.

“We’ve always had tools,” the ads say, “tools that made us stronger, faster, more capable. But they never replaced the people who held them.”

AI doesn’t just threaten to replace human beings, it is designed to do so, Aguirre said.

“ I know this sounds alarmist and it’s very difficult to take in, but this has not happened by accident,” he said. “Our economic system has been pouring more money, more human thought into developing these AI systems than any other project in human history.”

Open AI, a major AI company, has brought in $13 billion in annual revenue. 

People who love the AI now may not realize what’s coming, and there may be a fair amount of wishful thinking, Aguirre said.

The ads, he said, are a call to action. The public has “to demand change from the government and [for] the government to take action,” he said.

Because at this point in the AI timeline, “the government is the only entity remaining that is going to be able to [achieve] a course correction.”

‘A bare minimum’

That correction, he said, in part requires treating the AI industry the way the government treats the pharmaceutical industry.

If a drug company puts makes a new drug, it has to undergo all sorts of testing and review to see, for example, if it increases thoughts of suicide. 

The drug’s benefits and risks are compared, weighed, monitored, and regulated. 

AI systems, however, which carry the same risks and side effects, are not put through the same federal or state safety standards.

“If you roll out an AI system tomorrow to several hundred million people, nobody knows whether it increases suicidal ideation,” Aguirre said. “There’s no testing, there’s no evaluation. Even if it happens internally, it’s done quietly. It’s not published, and there’s nobody outside the company who can say, ‘Oh no, sorry, that’s too much.”

That needs to change, he said.

“We need a safety and external testing regime before AI systems get deployed,” he said, “bare minimum.”

And the fixes are relatively simple. If the laws require developers to require AI systems to be accurate and safe, then they will be, he said. If the law requires developers to prevent AI from doing something dangerous or helping someone commit a crime, then the systems won’t do that, Aguirre said.

“Somehow the companies imagine that they are exempt and regulators so far have treated them as exempt from the same basic measures that we apply to all of others,” he added.

That regime would ideally be at the federal level, but the Trump administration is seeking to block states from regulating AI on their own, arguing that regulation will stifle growth. 

But guardrails are not about strangling innovation, Aguirre said, they are about protecting human values.

The effort to regulate AI is already late, Aguirre said, but that doesn’t mean it’s too late. 

“ The time [to act] was yesterday or last year, or five years ago,” he said, “but the second best time is right now.”

Companies are not going to regulate themselves, he said, there’s too much money to be made.

Selling human replacement to investors

AI companies are promising investors a future where paychecks and health insurance, among the biggest expenses a company has, are minimal concerns, Aguirre said.

“What they really want and what they’re selling to investors is, we’re going to allow you to replace a significant fraction of [your] workforce purely with AI systems,” he said.

That bottom line, he said, “is the only thing, honestly, that makes the economics make sense for the vast investments that are being made.”

What happens when suddenly not just some, but most, companies start laying off their workers because an AI system is cheaper, much faster and much more efficient? Some companies are already laying off workers as they increase spending on AI. 

“Very quickly people will start to see that they simply don’t have any economic value. If an AI system can do most of the things that they do, even if it doesn’t do them as well it will do them for cents per hour,” Aguirre said.

The momentum is growing and heading in one direction, but it’s not too late.

“[We’re] going to need government action. We’re gonna need regulation, we’re gonna need governance, we’re gonna need standards, and we’re gonna need people demanding it.”