New York Governor Kathy Hochul thinks AI should be ‘less evil’

extract

Today, I’m talking with New York Governor Kathy Hochul, and I’ll just warn you — this episode moves really fast. It’s an election year, after all, with a shocking amount of tech policy at stake, and Governor Hochul has taken strong positions on almost every major tech issue there is. New York Governor Kathy Hochul thinks AI should be ‘less evil’ The governor on 3D-printed guns, AI and data center backlash, and who protects kids online. If you buy something from a Verge link, Vox Media may earn a commission. See our ethics statement. For example, Meta just reached a settlement with dozens of states, including New York, which will restrict how teens use platforms like Instagram in very specific ways. Governor Hochul is a strong supporter of those restrictions and more, as you’ll hear. But those come with a cost — widespread age verification means adults will also have to show ID to use the internet, which will essentially make it impossible to be anonymous online. Verge subscribers, don’t forget you get exclusive access to ad-free Decoder wherever you get your podcasts. Head here. Not a subscriber? You can sign up here. So I asked about that, and if these new restrictions amount to Meta conceding in a lawsuit that these kinds of regulations on speech might not pass First Amendment scrutiny if the governor were to try and impose them in a law. I had similar questions about New York’s restrictions on ghost guns, which prevent 3D printers from printing gun parts. That law is already being challenged — not in courts, but by activists like Cody Wilson, who runs the largest database of 3D-printable gun files and who has started altering those files in a process he calls “Hochulization” in order to show the ban is technically toothless. I asked the governor about this, and, well, you’ll hear her challenge Cody right back. Oh, but there’s more: We also talked about surveillance and Flock cameras, and, of course, the politics of AI and data centers in an election year. Governor Hochul signed a one-year moratorium on data center construction in July, and I wanted to know what she’s looking for when that expires to make data centers a workable proposition. We also talked a little about Mayor Zohran Mamdani in New York City and, of course, her org chart for the state government. This one is a lot, I think you’re going to like it. Okay: New York Governor Kathy Hochul, here we go. This interview has been lightly edited for length and clarity. Governor Kathy Hochul, you are the governor of the state of New York! Welcome to Decoder. Thank you. Happy to be on. I have so much to talk to you about. We don’t have a ton of time, so we’re going to move fast. Are you up for it? I’m up. All right. I want to start with what I call the Decoder questions, lightning round style. I think the operation of your office and state government is really opaque to most people. So I ask everybody this question. How is the governor’s office structured? How did you make decisions about how to structure that office? How did you organize it? Governor at the top. Then there’s a leadership team, senior leadership team. There’s a secretary to the governor who is the top advisor. It sounds like the person taking notes at meetings, but the secretary of the governor is the most powerful non-elected position in the state. When I first became governor five years ago, I said, “I want the smartest, the best, the brightest people I could surround myself with.” I didn’t know a lot of these individuals, but they were highly recommended from having business experience, government experience, or technological experience. I said, “Bring me the smartest.” There’s a layer of senior team members. There’s one individual who oversees all the agencies. There are 45 different agencies for everything from the Department of Transportation to the environment to our housing development agency, economic development. It all trickles down to a pretty large bureaucracy. I’m the decision maker, but I take input from a lot of different levels of government before I make a final decision. But I’m a fairly quick decision-maker. Show me the facts and I can make a decision instantaneously. You took over in a complicated way from your predecessor and then you won reelection. Did you sit down and make an org chart of how you wanted your office to work? Yes, I did. How long did that take? I inherited one in a sense, but also early on I didn’t have a lot of tools. In fact, we had a hurricane in my first week. I didn’t know how to declare a disaster. I hadn’t hired that person yet, the head of emergency services. We were caught off-guard a little bit. The state still functioned, but I knew I needed to do more. I’m always in search of talent, smart people, and trying to rotate people in. Sometimes I start commissions like the Future Works Commission to bring in experts, not just from my administration but around the country, who understand the displacement of jobs because of AI: who’s most vulnerable, how we deal with that, and how we train people to be able to overcome those obstacles or to embrace AI as a possible career. I’m always finding ways to supplement my team with industry- or issue-specific experts. This question’s going to sound ridiculous, but I’ve asked a lot of people a variation of this question. Almost no one ever is going to sit down and say, “I need to write a new org chart for the governor’s office of New York.” Did you do that on a whiteboard? Did you open PowerPoint? Was it just you alone with a laptop? What was it actually like? Can I tell you something I’m embarrassed of, since you phrased it that way? I did it on a legal pad. Nice. Old school. Sorry. Yeah, pen and paper. No, it’s great. Very few people do that. You would be surprised at some of the answers we’ve gotten to questions like that. I just laid out, “This is what I need, this is the team I need.” Now that was five years ago. I’ve gotten a little better. I’d probably ask Claude or ChatGPT for some recommendations. Oh boy. We’re going to come to that. Don’t worry about it. You started answering this question, but this is the other Decoder question. How do you make decisions? What’s your framework? Bring in all the information as quickly as possible, pressure-test it, and then go with my gut. One of the more unique parts about structure in New York government is the relationship between the state and the city. We’re talking here in person in New York City. Mayor Zohran Mamdani has been pretty aggressive about going after billionaires and tech companies. Most recently the city went after Amazon for its labor practices. Do you think that’s been effective here? I need to make sure that New York City and the state overall continue to be a place where we support innovation, new jobs. Technology is leading the forefront right now. I’m really proud that we now have been named the number-one job creator for tech jobs in the country, surpassing California, which has always dominated in this space. That says something about our ecosystem and I want to foster that. I want people who have just an idea who can use technology to create something that no one else had ever conceived of and be successful in marketing it. That is what I’m leaning hard into. The mayor has his own approach, but I think overall we both want New York City to succeed. To me, success also means creating more opportunities for jobs because of the spinoff effect. Five guys in an office right in this building could lead to 500 jobs in a couple of years. Those opportunities help lift people out of their circumstances and achieve what I’m hoping for is more opportunity for more people. There’s a WeWork just above us. We got to go find those five guys. I’m asking specifically about the more aggressive approach the mayor has taken. He’s taken a very aggressive approach. And like you said, the numbers show that the billionaires didn’t run, they’re actually investing; the tech companies didn’t run, they’re actually hiring at a rate that surpasses San Francisco. Is that emboldening you to take a harsher regulatory approach to tech companies? I don’t just take a regulatory approach for the sake of doing it. I don’t raise taxes for the sake of doing it. I do things in a smart way when it’s necessary. For example, we have regulations on tech companies when it comes to social media for young people under the age of 18. My job is to protect the young people of our state because they have been spiraling with mental health challenges — exacerbated by the pandemic, yes, but many of them have not overcome that sense of doom and anxiety. It is pervasive in our teenagers particularly. I look at my role to regulate tech companies in that space. Before the Meta settlement, we had innovated and come up with probably three years’ worth of reforms to force the companies to do things that they are otherwise not doing on their own, and one would wish they would but that’s probably being too optimistic. We have in place, for example, no notifications to someone under the age of 18 between midnight and 6:00 AM. These kids need to sleep. We’re saying addictive algorithms can no longer bombard young people, having — this is a big deal — automatic privacy controls that are already set. Parents aren’t expected to understand every platform where their kids may go or any gaming site and have to get ahead of them. That it’s automatic using age verification is really important. Limiting access to AI chatbots for young people and making sure there are disclaimers so they understand this is not a real person that you think you’re having a relationship with: that’s an example where I think tech needs to be regulated. It went out of control and now the government has to perform a function to protect our residents and our young people, the most vulnerable. I don’t just regulate to stifle their growth. I want them to grow. It’s a smart, strategic, more surgical approach to regulations. I’m glad you brought up the Meta settlement. That’s obviously the big news this week. Meta settled with 47 states, including New York. I look at that settlement, I look at the list of controls that Meta has agreed to and my initial reaction was this is a bunch of stuff that, if passed as a law, would run into the teeth of the First Amendment. They are effectively speech restrictions, right? We’re going to prevent young people from using these platforms to speak at certain hours of the day. And maybe there’s some argument that overcomes the First Amendment objection, right? The teen health risk is so high that this is actually a narrowly crafted, strict scrutiny-abiding restriction, but we don’t know that. That actually hasn’t been challenged in court. Meta just gave that to you. Then there’s age verification to prove that you’re over 18. Even if you’re an adult, you still have to show the internet your ID, which is a big ask for people who are not covered by the restrictions. That’s a big swirl. Do you think you got concessions from Meta that you otherwise would not have been able to get if you just passed the law? No, we actually have laws in New York that are more restrictive than some of what Meta has agreed to. Are you worried about the First Amendment implications of some of those laws? No, I’m not. And here’s why: we have an opportunity and the responsibility to protect our residents. The First Amendment is when the government says that we’re limiting your speech. Not having notifications that keep you awake all night when you’re a 13-year-old is not restricting your First Amendment right to free speech. There is a difference. Sometimes the First Amendment is used as any kind of government regulation. People will say, “Oh my God, you’re violating the First Amendment” or any of the other amendments. I’m an attorney. I look at it closely. I don’t do things to violate amendments or any laws. It’s not my practice. We’re smart about how we structure these and make sure that we’re protecting First Amendment rights as well as protecting young people from having their mental health monetized by companies who have a profit motive. It’s almost like cigarette smoking. The earlier you get them, the longer you’re going to keep them hooked. Young people should not be bombarded with the addictive algorithms or be subjected to the bullying that goes on. Someone who’s thinking about suicide searches for “suicide” and is now told different ways to commit suicide. That is terrifying for parents. Parents actually appreciate the fact that they have a voice in government. Maybe it’s because I’m the first mom governor in New York that I understand this. I’ve raised kids. I know the pressures they’re under. Something else we did: I’m really proud of the fact that we’re the biggest state in the nation to have come up with a full ban on cell phones in schools. Now you can say it’s violating their First Amendment rights. I’m saying I’m protecting them from a lifetime of addiction to a device where they no longer have human relationships. You asked how I make decisions. Sometimes I go with my gut, sometimes I want confirmation from the real world and I go out there and talk to people in schools and in cafeterias and have a lot of conversations with young people. When I did round tables around the state of New York for about a solid year to get information, [students] were so anxious about missing a notification. “The girls are meeting in the restroom to plan activities for Friday night and I just got excluded and I’m also getting knocked out because of what I’m wearing. And now they’ve got a picture of me in an inappropriate way being sent around the school.” These kids are paralyzed. One young person, probably a sixth grader, said to me, “You’ve got to save us from ourselves. We can’t put these down. We’ll be ostracized.” That’s when I took it to heart and I said, “I will protect you. I’ll protect these kids.” We literally instituted this in September. The survey results from just this past May show anywhere from 75% to 80% of the schools surveyed have been overwhelmingly positive. The kids who resisted at first are now making friends in person. There’s more communication. They’re learning the interpersonal skills that you’d want your future workforce to have, and they’re playing games at lunchtime. They’re playing cards. They have baseball cards they’re trading. They’re playing chess, bingo. It’s almost silly, but they’re just being kids again. And so they’ve got plenty of time for the stresses of the real world. They can become hooked on social media later on in life. They can get hooked on their device. They can do all this later, but why do we rob their childhood from them? Sure. I’ve got kids in New York public schools. I’m reasonably happy about the bans. I’m not excited about my daughter ever getting a phone. But if you were to do this straightforwardly in a legislative way, particularly the restrictions that Meta has agreed to, you would have to build out some record and you would have to do some balance of harms analysis. Here, Meta agreeing to age verification means that I, an adult, have to tell them how old I am with some tool, probably by showing them my ID. That is a big burden on me, an adult, that has nothing to do with the safety of the kids. That to me feels like it might not pass the scrutiny that the courts would provide if Meta was like, “Actually, this is too much of a burden.” How do you address that? Because the big concern here is that you’ve actually, instead of saying age verification for teens, de-anonymized the internet. The proof is the fact that Meta agreed to this in a settlement. Meta’s looking out for its corporate interest. I’m looking out for my interests as a citizen. Meta’s looking out for its ability to get more citizens like you to be engaged with it so it can make more money off of you and sell you more products. It’s in Meta’s interest to have you engage. The fact that they’re willing to step back— Right. They don’t care if I have to de-anonymize myself to look at Instagram ads for things I don’t need to buy. I’m saying I would like to be anonymous on the internet and if we put age verification everywhere, which is one of the consequences of this agreement, that every platform is going to have to start doing age verification at scale. I won’t be able to use the internet without constantly identifying myself to prove my age. I’m sure there’s a path forward that doesn’t require you to do this constantly. I’m not aware that you’d have to do it every single time you participate, every time you go to a site. I’m not aware that you’d have to do it that way. Let me find that out. I’m not going to say something without knowing that answer, but it’s an interesting point you raise. It really is. But I’m focused on the fact that we did gather data around the country to find out what is happening to our young people today. I cannot rely on parents to know how to stop this from happening, because the parents are the ones telling me, “Please do something. I don’t know how to stop this. My kid is spiraling.” One mom told me that she has her husband leave work every day at three to make sure their son, who’s being bullied in school and harassed and going to dark sites, doesn’t commit suicide when he goes home. Who wants to live with that specter? That was not the case when I was a kid, or my kids were kids. The world has changed and again, it is not a no to progress or innovation. I have thousands of examples of me inviting tech companies here. Even Micron, the largest private-sector investment in American history. 50,000 jobs are happening up in Syracuse because I went after them aggressively. They’ll be open in two years. This is extraordinary, building the memory chips that are necessary to power AI in New York state. My credentials are impeccable when it comes to inviting innovation in, but when there are areas where we need guardrails to protect individuals, I’m going to step in and do this. I feel good about this. This is a decision that we’ve made and I think it’s overwhelmingly supported by many parents. Young people someday will appreciate the fact that they were spared from this, and other people who adapt to this, this is the way it is. So much tech regulation, particularly for the last decade, has run into the First Amendment. Again, my view of it is we’ve avoided the First Amendment conversation by talking about Section 230 or other red herrings that allow us to not talk about the fact that a lot of this looks like speech regulation and expression regulation in important ways. Another good example of this, which I think will be developed, is the restriction on ghost guns in New York state. The idea is that you should not be able to 3D-print gun parts in New York state. We have a story just this week about Cody Wilson, who printed the first 3D-printed gun. He’s put out a tool to defeat the way that the law is designed to work. He calls it “Hochulization,” where he’s going to de-fingerprint the hashes of 3D-printed gun files so that whatever tool is scanning at the 3D printer level will not see that it’s a gun and print it anyway. That’s a cat and mouse game. That is a technological arms race. You can tell Cody I will get ahead of him next. I can’t always be on the forefront when I see someone intentionally thwarting a law. It is the law that you cannot have 3D-printed guns in the state of New York. And if you think it’s okay to have someone in their kitchen who may have severe mental health problems who wants to shoot up kids in your school, make a ghost gun at their kitchen table or create a munitions factory where there was none, with no regulations, no background checks, no nothing, and people broadcast that as if it’s a proper thing to do is challenge me on that, then I say go for it. Because you’re trampling in the wrong space. The question I’m asking is about implementation. I understand the desire to prevent guns from being 3D-printed and I understand that probably a lot of people would prefer that not happen. Yeah, it’s bad. It sucks. It’s wrong. But implementing that means at some point the printer itself has to know “this thing is a gun and I’m not allowed to do this.” That’s right. That is a kind of regulation on expression. We have a case for an open source e-reader that looks like a gun that people have made just to see if it will trip the software detector. Even though it’s not a gun, it looks like a gun, and there’s some line in there that is very complicated for a government regulation to address. How do you view that? Do you want the printers to literally look at things that look like guns or do you want it to be more technical, more like a database or a registry? I want to make sure that we’re getting the objective I’m looking for and I’ll figure it out. As I said, I assemble the smartest people when I make decisions, especially in the area of technology where the typical government employee is not proficient. So I’m going to use the foundation I have, but then go out there and say, “Tell me how to make sure that no one can thwart what I’m trying to do here,” and I’ll get it done. If we miss the first time, I’ll do it the second time and I’ll do it the third time because I’m not going to allow this to happen. Yes, there are other restrictions that occur. I get that. It’s not ideal, but I also didn’t think it was a smart idea for people to manufacture weapons of mass destruction at their kitchen table and kill kids. So I’m going to be the one who says no. I’d rather have the regulation that protects and prevents it. Use 3D printers for any kind of good. You can make all kinds of things with 3D printers. It’s not a ban on 3D printing. It is a use that is illegal because you cannot have a gun in the state of New York without having had a background check, having regulation, having it registered. We have all kinds of rules, which is one of the reasons why we are one of the safest states in America. Lowest gun homicide deaths in America. It’s extraordinary, because the rules are working. So not everyone loves rules. I get it. This is not the wild west. This is New York. We’re going to do it right. If there’s a theme emerging in this conversation, it’s how much can the government watch what happens on my computer and how much can it tell me what I can do with my computer? Which is maybe a 40-year-old conversation, but it’s more and more relevant over time. We’re saying we’re going to restrict how teenagers can use their phones. We’re going to restrict what you can print on a 3D printer. The flip side of that is... I would say the backlash to surveillance is bipartisan. The backlash to Flock cameras is deeply bipartisan. It’s scrambling party lines. Right. You’ve allocated budget for police departments to modernize, to install Flock cameras. How are you viewing that backlash to surveillance? I understand the anxiety around this. I truly do. But you see how it’s all the same, right? I do not have complete regulation over someone’s use of their computer. That’s an overstatement. It just is. We have a very narrow, specific purpose and a government objective to protect young people, because the companies have gone too far. If they had put these controls in place themselves, when the data was starting to show them that this is harmful to kids, [but they didn’t]. They stopped letting their own kids use this, but let everybody else be subjected to this, knowing the harms. That’s when the government has to step in and regulate business. We don’t do it as a first position. My view is more hands off. Innovate, do good things, sell your product. Great. This is called capitalism in America. A 3D printer, those are great. Until I find out they’re making guns that can kill people from them or be able to convert a Glock. Maybe it’s a legal Glock that you have registered to you, that’s great, but you can get this little piece of plastic that’s the size of a Lego and insert that and all of a sudden it can fire countless rounds per minute. Now it’s a weapon of mass destruction. The government has to be watching to an extent. Watch for excesses, but not to stifle personal use, not to stifle innovation, not to stifle the companies. When they go too far or they don’t self-regulate, using data that they have internally and ignoring it, thinking they can get away with it, that’s where we are today. I’m just connecting the dots between “I want to use my printer in some way,” “there’s a connection to the government that says yes or no,” and how people feel about technology like Flock — “I want to drive my car too fast on the highway and an automated license plate reader is going to track that and send me a ticket in the mail.” Or “cops are going to use this in ways, good and bad, and sometimes they’re going to go over the line and there’s not enough regulation,” or “I don’t want these cameras at all.” Which is what we hear from so many people in so many communities, across party lines. Actually let me read you a quote — To the extent that they’re being abused, that has to stop. Sure. The CEO of Flock recently said to a lot of people, including The Verge, that he was wrong to assume that the regulatory process would provide protections for civil liberties. He assumed that state and local governments would provide controls for Flock to prevent things like cops over-searching the database or tracking people. Tracking a lot of ex-girlfriends, for example, is a thing that we hear about cops using Flock for. That’s a pretty stunning indictment. The state is rolling out cameras to track people without the necessary controls, and the company providing the camera is saying, “Well actually you were wrong about this.” The state is not rolling these out. I give money for all kinds of technology. The localities are the ones that decide how they want that spent. I do not have a contract with Flock. That is a mischaracterization. But the money provided does not come with controls? I use the money for overtime for police. Computers in police vehicles is primarily the main focus of this. And to the extent that there are license plate readers, that’s different from the Flock cameras which are automatically transmitting information without human intervention, which is what I object to. Okay. If there’s a police officer that says, “This person is wanted for murder, let’s go get them,” you’d be happy if we went after them. But this is not just to have AI or a machine make an automatic referral of information to other agencies, whether it’s ICE or anybody else. That’s not what we’re talking about here in the state of New York. That’s not what we want to happen. If I need to tighten up regulations in that space, I will. I have no trouble looking— Can you connect that to the money that you’re allocating for this technology and say, “You have to use it this way?” Yes, we can. We can look at it. That’s an observation. That’s new information to me, that we should be having restrictions on and how they’re using it. Absolutely. What would you ask of the CEO of Flock in order to make that job easier? I’d tell him to make sure that the information can only be amassed a certain way. It has to be reviewed with human eyes to make a decision on whether or not this has to be forwarded to another law enforcement agency, the FBI or someone’s wanted list. That’s basically the purpose we want it for. I don’t want Big Brother. I don’t want unfettered surveillance of Americans that makes them anxious. It goes against the sense of liberty that we all have and deserve to have living in this country. This is not a country where the government should be monitoring everything you do, but to the extent that we have to stop crime and thwart the criminals who are always trying to take steps ahead of us, I have to get ahead of that as well, and try to manage that or else we have a lawless society. People are getting harmed and then it’s a failure of government to intercede. We’ll come to data centers, but I think data centers are doing this as well. There is some kind of realignment happening. These are not right- or left-coded issues. They’re not Democratic or Republican issues. They are scrambling that traditional very, very strong party ties that have existed for almost a decade in this country. Surveillance is one, data centers are another. Do you see that — that there’s a scrambling here? There’s nothing wrong with that. I welcome that. I have been waiting a long time for there to be more issues that people, all Americans, can galvanize behind and find some commonality. When I’m doing roundtables around the state, and I’ve done many, I don’t know if I’m talking to Democrats or Republicans. All I know is these are people worried about what’s happening in their communities with hyperscale data centers being built, making noise, causing disruption, and consuming energy. Is that driving up their own utility bills? They’re anxious about this. No, this is important. It’s an important time in our history to find that there are common interests and maybe that’s a breakthrough if you can have elected officials remind people like that: “Listen, you’re agreeing with your Republican neighbor and you’re a Democrat or vice versa.” That’s good. Greg Abbott, the governor of Texas, followed our leadership on the moratorium on hyperscale data centers — not the smaller ones for university research or educational institutions or healthcare institutions. We are very specific and these are the large-scale data centers. But to have the state of Texas follow our lead, that says something. That says something — that Americans overall are saying enough is enough. The rudest version of this question is: Doesn’t that just say that it’s an election year? No. You don’t think so? No. This only became an issue fairly recently in the state of New York. We were recently bombarded with 30 applicatio