Once again, our views are being manipulated—this time around AI. Highlighting some recent news that didn't get enough attention.
The New York Times recently reported that China, Russia, and to a lesser extent Iran have been working to fan the flames of the American debate over AI data centers. The article cites Alethea, a threat intelligence firm, as finding that state media in those three countries mentioned data centers roughly 700 times between January and June — nearly four times a day. These actors also made AI-generated propaganda easily shareable online. For example, in one instance, an actor in China used ChatGPT to generate a comic strip blaming data centers for soaring electric bills and dressed it up to look like the work of a reliable Maryland news outlet. The goal, as the New York Times reported, is to turn our data center controversy into a domestic fracture point. As Jessica Brandt, a former official with the Office of the Director of National Intelligence, told the Times, foreign actors aren't manufacturing America’s debates over AI, but “they are exploiting them.” The debate is geared to deepen our divisions and weaken us from the inside.
Reading the article, I realized something unsettling. I'd seen at least one of the posts it described. At one point, I might've even been tempted to share it.
Most of us know by now that social media warps perspective. We know the algorithms serve up the juiciest, most inflammatory posts — the ones best built to keep us glued to the platform while the ads roll in. We have fancy terms for it: “the attention economy,” “persuasive design,” and so on. And whatever our politics, most of us have thought, at one point or another, that the “other” side has been duped by propaganda fed to them by these algorithms. And yet we stay.
Worse, we know the algorithms make us hot. Fired up. Angry. Anger holds attention better than anything else now. Honestly, at this point, anger sells better than sex.
And today, people are angry about AI. I know, because I was one of them.
I knew all about the evil algorithms, but frankly (stupidly), I considered myself immune. Big brain, lots of education, analytical mind, blah blah, stuff that makes me feel smart. So when provocative posts about the evils of AI started filling my feeds — the data centers hogging water and making people sick, the distasteful tech elite, the jobs destroyed — I didn’t distrust them. I joined my friends in reposting sarcastic memes and bombastic statistics. Almost daily, I found something new to be pissed off about.
And, to be honest, I'd barely used AI at all. I tried it once, years ago, and found it lacking: robotic, prone to hallucinations, not very helpful. That was the end of the story for me. I knew (or thought I knew) that AI was bad, and I found it useless, so clearly I had to rail against it.
I started to change my mind when Google began forcing its AI answers on my searches. If AI was going to be shoved on me anyway, I figured I should at least find out how bad the water consumption and energy use really were — and whether Google was the AI I wanted to be stuck with. So I finally did the research I probably should've done from the start. And lo and behold: the issues turned out to be far more nuanced and complicated than my social media feed had let on. That’s a big topic, and a post for another day. The point is that I learned, once again, the hard lesson: social media and echo chambers make it a very susceptible time to be misled.
And the people doing the manipulating know it. It doesn’t matter who is farming your anger, foreign or domestic. The technique is identical: find your existing grievance, stoke it, fan the flames, blow the fire in their preferred direction, and watch it burn.
Consider what just happened in Manhattan. Alex Bores, a member of the New York State Assembly, ran for Congress in New York’s 12th District after having successfully co-authored the RAISE Act — one of the first and strongest state AI safety and transparency laws in the country. A super PAC network called Leading the Future — backed by the venture firm Andreessen Horowitz, OpenAI president Greg Brockman, the AI company Perplexity, and Palantir co-founder Joe Lonsdale — spent roughly $8 million to stop Bores.
Notably, the attack ads didn’t say a word about regulating AI. Instead, they went after Bores for his time working at Palantir, the Peter Thiel co-founded data analytics company that’s been under some political heat lately, particularly with respect to Palantir’s involvement with ICE. (Bores claims he never worked on the ICE contract and quit in 2019 when the company renewed it).
But wait a minute. Why would a PAC funded in part by a Palantir co-founder run ads attacking a candidate for having worked at Palantir? How does that make sense?
Because the PAC counted on the expectation that people weren’t paying attention to who was behind the attacks. Anger about ICE and Palantir was lying around, ready to be picked up and aimed. So the PAC capitalized on it, likely doubting many people would ever check the speaker. As another former Palantir employee told Fast Company: “If Bores’s campaign is one that would restrict the tech industry’s growth, and his base is one that is already primed to be critical of Palantir, people (like me!) who watch this ad wouldn’t suspect that it’s people with significant interests in Palantir and the broader industry that are funding the ads, too.”
To be fair, money flooded in on both sides. PACs favoring AI regulation eventually spent more than $19 million supporting Bores, over $13 million of it from a network indirectly funded by Anthropic, the AI company behind Claude. The pro-regulation side outspent the attackers two to one. Bores still lost the primary. But, to his credit, the man who beat him, Micah Lasher, a co-sponsor of the RAISE Act himself, used his victory speech to tell the AI companies he wouldn’t be taking cues from them. So the seat did not go to an industry ally, and the ads may not even have moved the result.
But the result may not have been the point. One political strategist told Politico the goal was “to scare other legislators into submission.” The message was aimed past the voters of NY-12, toward instead every politician thinking about writing the next RAISE Act.
Here is what I keep coming back to. Online, the fight looks like AI versus no AI. Stop or go. A binary set of choices. But follow the money and you find a different fight: regulated AI versus unregulated AI. Almost no one is seriously funding “shut it all down.” But plenty of powerful, well-resourced people are massively funding “no rules and regulations” — which is a problem, because the no-rules side does not need to win the argument about regulation: it just needs to make sure we never have it. Every hour we spend raging about whether AI should exist is an hour we do not spend on the questions that will actually be decided in the next few years: by whom, with what oversight, with what accountability, at whose expense. If they can keep us fighting about anything other than the scope and details of AI regulation, they win by default.
This is the distortion in our attention. If the ad attacking a company is funded by that company’s co-founder, something is off. If a post gets your blood pounding and your hackles raised, ask who benefits from where your anger lands. If everything you hear is “AI is bad in all ways,” you are being handed a too-easy answer. The real debates — water, energy, jobs, safety, power — are complicated. They deserve better than our reflexes.
Because rarely is anything as simple as a binary. AI needs our attention right now. Our deep, real, sustained, careful attention. Not the kind an algorithm can harvest. The kind we choose to give.