Skip to content

Podcast

The Rise of the Anti-Ai Movement

The AI Daily Brief: Artificial Intelligence News and Analysis

Source ↗ ← All highlights
  • Widespread Public Skepticism In The U.S.
    • Public skepticism toward AI in the U.S. is widespread and measurable in polls.
    • Multiple surveys show most Americans distrust AI and expect negative economic effects. Transcript: Nathaniel Whittemore And to be honest, calling it a movement might be a little overstated at the moment. It is certainly not one big organized thing. The reasons for it are not monolithic. And it would be reasonable for one to ask, to what extent this is just driven by media narrative. It is certainly the case that for the last few years, since the launch of ChatGPT, we’ve had a pretty never-ending cycle of AI hype and enthusiasm as the main media story around it, followed By some type of skeptical narrative, when that excited narrative fails to be exciting anymore. And the media is all over this one right now. Just this weekend from the New York Times, we got a piece called People Love the Dotcom Boom, The AI Boom? Not So Much. In it, author David Streitfield asks, Tech leaders are beginning to worry about the public’s underwhelming enthusiasm for their plans to remake the world with artificial intelligence. Will that burst the bubble? And then of course, even more commented on the Time magazine cover from last week called The People vs. AI. The cover features nine headshots of people who, for various reasons, find themselves opposed to AI at least as it’s currently constituted, forming the basis for a large cover story About this emergent political and societal force. For some, they believe that this is all somewhat deserving of scorn. Lumps on X shared a remixed version of the Time magazine cover, this one featuring nine cave people, with the headline being The People vs. The Wheel. Grunts from the silent majority dragging rocks and rolling as a drag. I’ve also seen plenty of versions of this with The People vs. The Internet or The People vs. Any Other Technology. Now, of course, what’s being implied here is that with any new technology, there is always some amount of resistance that eventually looks silly and short-sighted in retrospect. And for those who are most involved in this technology, I do understand the frustration of feeling assailed for building or working on a thing that you think is going to be really positive, But where it feels like so many people are genuinely mad at you just for doing what you’re doing. And yet, I think it would be a mistake to view the rise of anti-AI sentiment as simply a media narrative. There is a huge and growing canon of studies that show particularly Americans have extreme skepticism around AI. A recent YouGov study found that 58% of Americans said that they don’t have trust in AI versus 35% who do. 45% of Americans said that they think that AI’s effect on the economy will be mostly negative, versus just 16% who think that it will be more positive than negative. And nearly two-thirds of Americans in that YouGov poll, 63%, think AI will lead to a decrease in the number of jobs available in the U.S., versus just 7% who think that it will increase The number of jobs. In a Pew Research poll from last year, the U.S. Ranked dead last in terms of the ratio of citizens who were more concerned versus more excited about AI. Only 10% of those polled said that they were more excited than concerned versus 50% who said they were more concerned than excited. (Time 0:01:33)
  • Data Center Protests Signal Local Power
    • Local protests against data centers show anti-AI sentiment can produce immediate political wins.
    • These events signal broader public doubt about whether AI will benefit communities. Transcript: Nathaniel Whittemore And what’s more, it seems to be growing. A number of videos out of New Brunswick, New Jersey went viral on X after hundreds of citizens showed up to a planning meeting about a data center and got the project canceled. That video shared just a couple of days ago got 5 million views. Organizer Ben Zobiak, who posted the video, wrote, A data center in New Brunswick was canceled tonight when hundreds of residents showed up. When we fight big tech and private equity, we win. AI curator Andrew Curran writes, After three years, it seems to me that public anti-AI sentiment in the West is now at its highest point. Political commentator and statistician Nate Silver has been talking about this a lot. He recently tweeted, If AI produces unprecedented levels of technological disruption on timescales that are an order of magnitude or two faster than anything in human history, it’s Going to be an unprecedented political fight. And for what it’s worth, the timelines potentially line up with the 2028 US election. He also talks about how this white-collar disruption doesn’t have political precedent. He writes, Usually these transitions would take decades and not affect white-collar workers first, who have more political power. Displacement of narrow classes of blue-collar workers, e.g. Coal miners, even if most of society would benefit, usually causes huge political battles. Here, the displacement would be much broader, the benefits are less clear, and the people displaced are more influential. In another tweet more specifically focused on the data center concern, Nate writes, Opposition to building data centers might be irrational at the micro scale, they’re just going To be built somewhere else, but at the meso scale, people are profoundly doubtful about whether AI will broadly benefit society, and that’s not so irrational at all. (Time 0:04:26)
  • Engage Widely And Acknowledge Concerns
    • Engage with critics and acknowledge legitimate concerns rather than dismissing them.
    • Builders must talk to broad publics if AI truly impacts everyone. Transcript: Nathaniel Whittemore In fact, I think the leaders of the AI industry have done a spectacularly bad job of both acknowledging and addressing real concerns that people have. We may want things that are exciting to us to be exciting to others. To have things that are self-evident or obvious to us be self-evident and obvious to others. But that’s not how the world works. And if we acknowledge and believe that the impact scale of this technology is going to be circa everyone, we have to be willing to engage with circa everyone. (Time 0:06:47)
  • Anti-AI Is Many Movements, Not One
    • The anti-AI space splits into distinct camps like safety advocates and capability skeptics.
    • Different groups hold divergent premises, so the opposition is not monolithic. Transcript: Nathaniel Whittemore We’ll start first with the AI safety folks, the people who are concerned about X-risk or existential risk, who talk to each other about what their P-doom is. P-doom being probability of doom, that AI ends us all. These are the voices that you will sometimes find AI folks critiquing as living in the realm of the sci-fi. It’s the paperclip maximalists who argue for why superintelligent AI would be likely to end humanity, not even necessarily out of malice, but because we stood in the way of accomplishing Some goal that they were trying to achieve. The voice of this group was much louder right after ChatGPT was launched, probably for the reason that you could get crazy headlines about how the robots were going to kill us all, and Folks like Eliezer Yudkowsky to come in your magazine and write an essay to that effect. Now, just like within the anti-AI movement overall, and frankly on the opposite side of the AI booster train, there is a massive spectrum of folks even just within this concern range. Many of them operate from a place of genuinely good faith, which creates much more room for discussion even if you fundamentally disagree with everything that they think. You can go check out the AI Safety Memes account on Twitter as an example, or even check out Laurent Shapira’s Doom Debates podcast. I think even if one fundamentally disagrees with the concerns animating these folks, it’s worth having a bit of epistemic humility about all of this, given that none of us know the actual Future. Now, an interesting note about these folks is that unlike some of the others in the anti-AI space, they actually quite agree with the accelerationists on how powerful AI is. It’s just they’re very concerned about what the implications of that are. Interestingly, I don’t really think that these folks have been as much of a driver of the conversation as it might have seemed if you looked just in the first six months after ChatGPT Was launched. Going back to that tweet by Andrew Curran about how public anti-AI sentiment is now at its highest point, he continued, the primary driver by far is not X-risk, but concerns about employment And the impact on art. Safety advocacy might have been more effective and might now be in a much stronger position if they had emphasized societal and economic impacts more than X-risk over the last few years. (Time 0:08:17)
  • Safety Advocates Vary But Acknowledge Power
    • AI safety advocates share concerns about existential risk but vary widely in tone and emphasis.
    • They often agree with accelerationists on capability while differing on implications. Transcript: Nathaniel Whittemore P-doom being probability of doom, that AI ends us all. These are the voices that you will sometimes find AI folks critiquing as living in the realm of the sci-fi. It’s the paperclip maximalists who argue for why superintelligent AI would be likely to end humanity, not even necessarily out of malice, but because we stood in the way of accomplishing Some goal that they were trying to achieve. The voice of this group was much louder right after ChatGPT was launched, probably for the reason that you could get crazy headlines about how the robots were going to kill us all, and Folks like Eliezer Yudkowsky to come in your magazine and write an essay to that effect. Now, just like within the anti-AI movement overall, and frankly on the opposite side of the AI booster train, there is a massive spectrum of folks even just within this concern range. Many of them operate from a place of genuinely good faith, which creates much more room for discussion even if you fundamentally disagree with everything that they think. You can go check out the AI Safety Memes account on Twitter as an example, or even check out Laurent Shapira’s Doom Debates podcast. I think even if one fundamentally disagrees with the concerns animating these folks, it’s worth having a bit of epistemic humility about all of this, given that none of us know the actual Future. Now, an interesting note about these folks is that unlike some of the others in the anti-AI space, they actually quite agree with the accelerationists on how powerful AI is. (Time 0:08:26)
  • Capability Skeptics Can Hinder Adaptation
    • Capability skeptics downplay AI as ‘autocomplete’ and resist updating views after progress.
    • Whittemore argues they harm public readiness and individual economic adaptation. Transcript: Nathaniel Whittemore These are the folks you see running around on social networks claiming things like AI is just fancy autocomplete. Demonstrating just how much the anti-AI space is not one thing, these folks disagree with the safetiest because if AI is just fancy autocomplete, it’s obviously not going to take over The world and turn us all into paperclips. If you’re looking for an example of this, one of the most prominent examples is Gary Marcus. And honestly, believe it or not, this is the group that I kind of have the most frustration with. The people who’ve been lobbying these arguments take advantage of every time the media narrative shifts against AI to update their previous essay about why AI has plateaued, despite The fact that it’s plateaued at a point significantly more advanced than the last time they said the same thing. I think in general, the capability skeptics are pedantic, pretty viciously intellectually rigid and unable to change their perception based on new evidence, and ultimately, almost Entirely ignorable. I don’t think we should mistake the fact that a few people can make careers by being the token skeptics on the panel, with their perspective actually being useful. And the reason that I have the most frustration and animosity towards this group is that these are the ones who so many normal folks want to be right. They want them to be right so they can safely ignore this thing that they don’t particularly like until it fades like NFTs. I genuinely believe that many people will not engage meaningfully with AI because the capability skeptics aid and abet their natural disinclination to do so. And because of that, those people will be extremely far behind when it comes to AI usage. They will be the least adaptable to AI disruption, the least likely to benefit from new opportunities that AI creates. (Time 0:10:16)
  • Market Skeptics Versus Tech Skeptics
    • Multiple skeptic types include market bubble critics who accept AI’s power but distrust valuations.
    • Distinguishing tech skepticism from market skepticism clarifies debates. Transcript: Nathaniel Whittemore Next up, a group that has some overlap with the skeptics but are ultimately a different thing. These we’ll call the AI bubblers. They’re not necessarily or at least a priori skeptical of the capabilities of AI. In fact, many of them are convinced in the long-term power and disruption potential of AI. They’re skeptical of the business side of things. They’re skeptical of where value is going to get realized. They’re skeptical of whether today’s business models can support the deal and debt structures that are financing the industry. They’re skeptical of the market’s valuations of these companies. Maybe the most prominent of these, at least right now, when it comes to a media profile, is Michael Burry of Big Short fame. One of the things that I’ve always said on this show is that it’s extremely important to break apart skepticism of the technology itself from skepticism of the market’s interaction With the technology. It would be a completely coherent intellectual position, in other words, for someone to think that AI is going to and even is now radically changing things and still not think the market Is pricing the companies behind it correctly. The next category we’ll call the artist advocates. Now, in some cases, these are folks in the artistic or entertainment fields who are frustrated that AI is doing the things that they and their peers used to do. Others are concerned about copyright and IP more generally. And then, of course, a big group of these folks are kind of just normal folks who aren’t necessarily artists themselves, who just have a general uneasiness of the fairness of things, Which is not something that, for example, Supreme Court decisions about copyright and AI are going to solve. (Time 0:16:05)
  • AI Concerns Focus On Child Development
    • Concern about AI’s effect on children and teen development is prominent in religious and conservative communities.
    • Fears include chatbot dependency and weakened human relationships. Transcript: Nathaniel Whittemore They’re concerned about human relationship structures. They’re concerned about the impact of AI on child development. They’re concerned about what happens when people want to talk to AI girlfriends and boyfriends more than real-life girlfriends and boyfriends. I found that this category of concern is invisible to many who aren’t around the circles where it’s important, but incredibly pertinent, in fact, maybe the very top of the list, in certain Groups and communities. (Time 0:18:18)
  • Nurse Wins AI Protections In Contracts
    • Nurse Hannah Drummond helped negotiate AI protections into nursing contracts after seeing harms.
    • She advocates strict controls and testing before patient-facing AI is deployed. Transcript: Nathaniel Whittemore An example of this from the timepiece is nurse Hannah Drummond. Now, based on Time’s profile, which is of course all I know about Hannah, it doesn’t seem to me like she’s some ideologue who rejects AI on principle. It’s that she has specific concerns with how it’s used in her field, which she understandably feels like she has some ground-level lived experience insight into. Time writes, Drummond helped nurses at 17 facilities in the HCA hospital system, including her own, win AI protections in their most recent contract, including a provision requiring Hospitals to give registered nurses a say in how new technologies related to patient care are implemented. Drummond points to concerns that nurses have had where the improper use of AI has not only been not helpful but actively harmful. One example she gave was of an AI tool that was designed to automate shift handoffs that had assigned a patient with COVID alongside another who was immunocompromised, which created Obviously big risks for the immunocompromised patient. Drummond Time writes doesn’t want to ban AI from hospitals but says there need to be strict controls around its use. She says, everything that reaches patients in healthcare has gone through rigorous testing and has proven to be safe, effective, and free from harming us. Why would we cut out those same test points for this? (Time 0:20:04)
  • Big Tech Power Shapes Distrust
    • Big tech resentment fuels anti-AI sentiment tied to power concentration and political influence.
    • Historical scale of tech’s reach raises legitimate governance questions. Transcript: Nathaniel Whittemore There are a bunch of flavors of this. One flavor that’s increasingly popular in political rhetoric is the tech billionaires as partisan villains. This is, of course, more prominent, although not exclusively prominent, on the left, and is aided and abetted by tech’s loud shift towards Trump in the most recent election cycle, As well as the very visible involvement of people like Peter Thiel and Elon Musk on a national political stage, there are folks for whom hold aside any idea of Marc Andreessen as a supervillain, Just have concerns in general about how much power the big tech companies have accrued. And again, going back to the core idea at the heart of this episode of having humility even around positions that one doesn’t hold, it is absolutely the case that up until the last generation, We had never had companies that touched billions of people in the way that big tech companies do. That previous types of companies did. And however one wants to resolve them, there are lots of legitimate questions around the balance of powers between civil society, governments, and companies, with big tech being A quintessential example. (Time 0:21:16)
  • Social Media Backlash Colors AI Views
    • Disillusionment with social media’s effects colors public views of new technologies like AI.
    • Many who once welcomed tech now fear repeat harms rather than progress. Transcript: Nathaniel Whittemore Matthew Iglesias wrote about this recently, saying all discussions about AI happen in the shadow of the tremendous and very sincere optimism about the cultural impact of social media That existed 15 to 25 years ago. I think that even people who are very technology forward and progressive have massive concerns about the last generation of technology social media, which you can see in the way that Millennial parents are handling social media for their kids. I have a four, almost five-year and a seven-year and none of the parents in and around our community are even considering giving their kids smartphones when they’re young. We’ll see how this changes in practice, but right now, pretty much all the parents I know who are not, I will be clear, Luddite anti-technology folks, Many of them work in the technology Field, but not a one of them wants their 12 or 13-year interacting with Instagram. Indeed, when you zoom out, I think that there are a few ingredients that are creating a particularly caustic environment for AI right now in America. One, like I just mentioned, is that many are not convinced that we’re better off because of the internet, and so have a hard time accepting that all technology is inherently progress. (Time 0:22:33)
  • Use Careful, Human-Centered Leadership Language
    • Leaders should avoid dehumanizing language that frames people as inefficient compute.
    • Speak to how AI augments human life and show concrete societal benefits. Transcript: Nathaniel Whittemore Sam Altman recently said in a public discussion, people talk about how much energy it takes to train an AI model, but it also takes a lot of energy to train a human. It takes like 20 years of life and all the food you eat during that time before you get smart. Context engineer Maratzen Koylan responded to this, I build AI for a living. I believe in what we’re building. But this kind of rhetoric makes my work harder and more dangerous. Comparing human development to model training is tone deaf and strategically reckless. People are losing jobs. They’re getting angry. They’re seeing AI as an enemy instead of a solution. Some are planning to destroy data centers and the people who build this stuff. That anger and backlash might not be reaching your floor, but it reaches the engineers and builders doing the actual work. The CEO of the most visible AI company should not frame humans as inefficient compute units. Your role as a leader is to show how AI solves real problems for humanity, not to reduce human life to an energy accounting problem from a comfortable position. I’m a techno-optimist. I believe AI enhances human capability. I work with this new form of intelligence every day. I genuinely respect what it is. It is real, significant, and unlike anything that has existed before. But I also believe in human excellence. We have to accept that it’s two fundamentally different forms of intelligence working together. In my humble opinion, the real techno-optimist position isn’t AI is cheaper than humans. It’s we now have two forms of intelligence on this planet, and the combination is more powerful than either alone. Speaking to Sam directly, he concludes, you’re the leader of open AI, and whether you chose it or not, you represent everyone building an AI right now. (Time 0:25:00)
  • Data Center Issues Are Policy Solvable
    • Data center community issues like energy costs are solvable with better policy and planning.
    • Whittemore sees a failure of imagination by builders to make data centers community-positive. Transcript: Nathaniel Whittemore There is absolutely no reason, with the economic structures being what they are, that data centers couldn’t be some of the most pro-community, positively engaged types of businesses Wherever they operate. By the way, even Donald Trump is on this train, increasingly pushing the AI operators to make commitments to ensure that they are better stewards for the communities that their data Centers operate within. (Time 0:28:07)
  • A Window Exists To Shape AI Policy
    • Political discourse on AI is not yet polarized and policies are still forming.
    • This window offers opportunity to shape pragmatic, nonpartisan solutions. Transcript: Nathaniel Whittemore It hasn’t hardened from a partisan perspective, and even the policies that people want to address all these issues are still forming. And there’s still a ton of room to shape perception of what should be done. I’m sure that as political campaigns take on a bigger portion of Mindshare in 2026, we’ll be talking more about what some of those proposals look like. But the point is that right now, the world has not, despite what X might make you think, hardened into the transhumanist accelerationists on the one hand and the neoluddites on the other. The vast, vast majority of people sit in the middle, trying to make sense of what this is all going to mean for them, their families, and their communities. They are not going to accept blindly that somehow this is just going to be a good thing, but most of them are also not going to reject out of hand the possibility that it could be. (Time 0:28:33)