社交媒体禁令有效吗？+ 关于 AI 意识的对话 + 工具时间

Kevin, I thought this was interesting. Meta sometimes struggles with the amount of trust that users have in it. Have you noticed this?

I have, yes.

People don't always trust that when they say something, the company is going to do right by them. So they have this new, interesting approach that they're taking to build trust. I saw this in the Financial Times this week: Meta is now testing AI glasses that continuously record audio and take photos every few seconds.

Oh, good.

Yeah. So if you were worried that putting Meta glasses on your face wasn't going to contribute to building a global panopticon, rest assured they will now just be continuously recording.

I'm so glad they've learned some lessons from all of their privacy scandals, consent decrees, and settlements over the years. It's really nice to know that they've taken all that to heart and set out on a better course.

Yeah, I hope they call these new glasses Cambridge Optica. You know, the Cambridge Optica version of the Meta glasses. I'm excited.

Cambridge Analytica? No, we're—

Yes, Cambridge Analytica would be another—

We're—

Another approach.

We're workshopping over here. Call us, Mark.

Call us.

I'm Kevin Roose, a tech columnist at The New York Times.

I'm Casey Newton from Platformer.

And this is Hard Fork.

This week: Are social media bans for teens working? Conflicting new evidence from around the globe. Then NYU Professor Jeff Sebo joins us to discuss new research into whether AI could one day become conscious. And finally, it's show and tell in our latest edition of Tool Time.

And it's a cool time.

Well, Casey, it's time to check in on a story we've covered periodically on this show, which is the state of the social media backlash and these social media bans that have been going into effect in countries around the world.

1. Social Media Bans Go Global

Yes, Kevin, you may remember that at the end of last year I made a prediction that by the end of 2026, 16-plus would become the new norm for getting a social media account worldwide. As we enter July, Australia, Brazil, Indonesia, Malaysia, France, the United Kingdom, Denmark, and Slovenia have either enacted laws or are preparing measures that would limit children's use of social platforms, typically barring kids under 15 or 16 from TikTok, Instagram, Facebook, YouTube, and X. That's a big change.

That is a big change, and I remember you making that prediction. But I also remember that you didn't include Slovenia in your list of countries that would apply one of these bans, so I'm only going to grant you partial credit on that one.

Here's what I've always said about Slovenia: They could do anything. It's a very dynamic place, is what I'll say about Slovenia. That was the prediction, but over the last few weeks we have gotten several big updates that make me feel like, number 1, this prediction is absolutely going to happen. More importantly, I just think the internet is going to start feeling like a very different place sometime soon.

Okay, make your case. Maybe start close to home. What is happening in the U.S. with social media bans?

2. America Embraces Age Verification

On Monday, the U.S. Supreme Court declined to block a Texas law that requires app stores and developers to enforce age verification and make kids get parental consent to download apps. So this is one of the stricter forms of age assurance, as it's called, that you will find out there.

Texas passed a law in 2025 called the App Store Accountability Act. If you live in Texas and you're a kid, you have to link your account to a parent or guardian, and then your parent or guardian has to approve any app before you, as a minor, can download it.

So the Supreme Court did not take up this case, which basically means that—what? This is now legal to do? If you are a state, you can pass a law saying that you have to age-verify to get onto social media?

Yeah, there was a legal fight over it. Industry groups tried to fight it in court, and a judge at one point blocked the law on the grounds that it likely violated the First Amendment. But in June, the 5th U.S. Circuit Court of Appeals put that judge's order on hold.

When the Supreme Court says, "Hey, we're not even going to look at this," that means that, yes, it goes into effect in Texas, and it will likely give other states that want to do the same thing carte blanche to do so. In fact, there are 4 other states that have passed laws very similar to the one that Texas has passed.

Which ones? Name them from memory. Don't look at your laptop.

I would love to name them from memory. They are, of course, Utah, Louisiana, and Alabama. So I guess it was really 3 states, plus Texas.

Okay, very good.

But California, interestingly, has passed a similar law that's less restrictive. It takes effect on January 1 of next year, and it requires operating system providers to collect age or birth date information when you're setting up a new device.

What is interesting about that group of states is that now you have red and blue. This is becoming a bipartisan thing across the country.

Are all these states defining social media similarly? Are they all including the same websites? I know YouTube has been on the cusp for some advocates of these bills, because obviously there's educational stuff on YouTube, and kids are using it in school. But it's also social media, at least the way that some of these attempts to age-gate it have portrayed it.

Yeah, for a minute it did look like YouTube was going to be able to wriggle out of these restrictions, but increasingly it is being lumped in with these other apps. I think it's important to say that this doesn't mean that a kid will be denied from using YouTube, because of course you can still open up a website, type in YouTube.com, and play videos. But it will prevent you from using an account.

Advocates would say this does have safety benefits. For example, nobody can send a message to your kid if they don't have an account. Maybe the recommendations won't be perfectly tuned to their particular brain rot if they aren't logging in every time they watch YouTube.

Right. So those are the state-level bills. Is there any movement on a federal ban on social media? That was something that you and I had talked about earlier—the U.S. government actually stepping in and doing something like this nationwide. But are we seeing any movement on that?

Not really. Congress has been debating a raft of other child-safety measures and, as usual, it seems like it's getting very close to the finish line and then falling apart at the last second. We've seen that several times over the past few weeks.

I think what is interesting in the context of whether we'll ever see a federal ban is that Pew published new research last week which found that nearly 6 in 10 U.S. adults do support banning social media for under 16. This is now the majority position in the United States, and I would guess that if eventually you see, let's say, half to three-quarters of states implement a ban like this, then that's probably when Congress will feel confident enough to pass it at the federal level.

3. Australia Tests The Ban

The first kind of ban like this that we ever talked about was the one that went into effect in Australia—

Yeah—

—which I think has been the template or the model for a lot of these other bans going on around the world. My understanding is that we're now starting to get some data back about how the Australian social media ban is going. There was a very good story by Jon Herman in New York Magazine recently about how hard it's been to ban teens from social media in Australia. What are some of the details there? What do we know about how that experiment is playing out?

Yeah, so there was a new study examining the early effects of the ban that came out 2 weeks ago. Basically, these researchers at the University of Newcastle in Australia did a 90-day check-in, and the big headline was that it found more than 85% of kids reported using social media roughly 3 months after the ban took effect.

That led to a lot of people saying, "Aha, I told you so. These sorts of bans are impossible," or, "Aha, look at these terrible social media companies. They're doing nothing to kick kids off." Australia's government, in response, is now introducing legislation that would double the fines against social media companies that do not take this more seriously. But I have to say, Kevin, as I dug into it, I became a lot less convinced that this is a failure.

What do you mean?

When you look at what Australia's law actually requires, the requirement is that platforms take reasonable steps to try to keep kids off of these platforms. It does not dictate a particular method.

What we're seeing the platforms do is adopt a range of strategies, one of which is what's called age inference, which is looking for signals that you're underage. For example, if you create a new Snapchat account and everyone else you message seems to be a young teenager, this may give Snapchat a signal that you're underage.

Yeah, if you're using the word "low-key" a lot—

That might be a sign that you're a teenager.

No, the people who are saying “low-key” are now solidly in college, Kevin.

Okay. You need to update your references.

I'm sorry, that's the youngest slang I know. What are the teens saying now?

I don't actually know, and I'm proud of that.

Teens, tell us what you're saying—

Yeah, if—

—to get past the filters.

Yeah, what's in your lexicon? Email Hard Fork at nytimes.com. But the point of the story is that it takes a while to develop these signals, right? And so if this is going to be one of the main things that you're relying on, you're probably not going to have banned 90% of the teens on your platform within those 90 days.

It also just seems incredibly easy for teens to fool these age-verification systems. There was a detail in Jon Herman's story about how people on Reddit are talking about submitting black-and-white photos of Thomas Edison to these age-verification systems to fool them. So it seems like the systems themselves are not perfect.

Well, when you implement one of these bans, you have to decide how much of a hardass you're going to be. It's possible to be a real hardass. You know who's a real hardass? China. You know what they require to create an account online? An official government document.

A vial of blood.

I mean, getting close, right? And in these democracies, we're not seeing that yet. They're trying a more gentle approach, in part because they don't want it to be an enormous pain in the ass every time that you and I, as adults, want to do anything online, right? We don't want to have to upload our driver's license to create a YouTube account, right? But that is what they have to do in China.

So far, they've taken this gentle approach, and I think that they're hoping that a combination of methods will eventually start to filter these teens out. And I actually think that there's an argument that this is the right thing to do, that swinging the pendulum all the way over to upload official government ID to do everything would be terrible overreach, and these somewhat leakier methods, while less effective in the short term, are probably easier to handle.

But if the net result is that all the teens in Australia are still using social media even after they're technically banned from doing it, why are we doing any of this?

Well, we're doing this because we assume that over time teens are going to use social media less, that by increasing the amount of friction over time, you're just eventually going to make alternatives look more appealing. It's kind of like, once they got rid of Napster, teens didn't immediately stop downloading music illegally. They were on LimeWire and a bunch of alternative services.

But over time, that got harder and there was more malware in LimeWire and it seemed riskier to use, and then one day Spotify shows up and people think, “Well, I'm just going to buy a subscription.” So I think we're going to see something here that is similar, where it's going to become more and more of a pain for these teens to be on these networks, and so they're going to start to find alternatives.

Interesting.

Yeah.

4. The Teen Mental Health Debate

Now, Casey, I want to talk about something you wrote which was about this Candice Odgers character. She has emerged as the foremost critic of the Jonathan Haidt school of “Let's ban the phones in schools, let's ban social media for under 16.” She and Haidt have been at each other over this issue of whether this is even a good idea or not, and you seem to be taking Haidt's side and saying, “Well, I'm not sure about this research or these arguments that Candice Odgers is making.” So maybe run down her arguments and how they contrast with what you believe is going on.

Yeah, and let me first say I think Candice Odgers is a really thoughtful critic and a great researcher, and I have enjoyed reading her work on this subject.

That's a prelude to dragging her ass. Okay, go for it.

Well, I saw her give a TED Talk that was posted online recently. The actual talk took place in April. And I did take exception to some of the ideas in it. I think what Odgers says in the TED Talk is that these bans don't work and that they let tech companies off the hook, and so we shouldn't do them.

Among my criticisms of that idea is that we just don't have enough data to be able to say that they don't work. Her argument in the TED Talk is, “I've talked to teens. They tell me they don't work.” My argument against it would be, “Let's give it more than 90 days.”

Also, it's not really a principled argument to say we shouldn't do bans because bans don't work. If you think that the bans could work, you would just make the ban more effective.

Yeah.

Yeah.

I mean, one of the arguments that folks like Jonathan Haidt made, which we asked him about when he was on the show most recently, is not just that these bans would work and that they would get teens off social media, but that they would also improve the mental health or well-being of teens. So I know it's still early, but do we have any kind of evidence about whether it is actually improving teen mental health?

No, we don't. But I think it's important to point out that on the whole subject of social media bans, teens, and social media, there have been 2 eras of argument. The first era was, “Hmm, it seems like there's a mental health crisis involving teens. Let's try to go out and study teen mental health at the population level and try to see to what extent we think social media is playing a role in affecting it one way or the other.”

Typically, when people have gone out to do this, they've either found no effect or the effects are very small. And this is the Odgers argument: You're pointing the weapon at the wrong thing, okay? The innovation that Haidt and his collaborators make is that they come along and they start looking at all of the teens that have been groomed and sextorted and lured into danger and scammed, and they're realizing that millions of teens are making these reports every single year.

Those are direct harms being experienced by an enormous group of people. You don't need a sociologist to study population-level effect sizes to be able to confidently say that millions of teenagers are being harmed. And so what Haidt says is, “If you got the 13- and 14-year-olds off of social media, you would spare millions of them from having these terrible experiences.”

So that is the thing that the Odgers side of the argument doesn't reckon with at all, and I just think it's terrible that that's happening to all those kids.

Is her argument—and the argument of the folks in her camp—more that there are offsetting good things that come from having teens on social media that outweigh the harms? Or is it that the harms are just wrong or being misattributed—that these platforms aren't actually as bad for teens as we're saying?

Some people argue what you just argued, and Candice Odgers does not make that argument in her talk. Where she makes the argument is in saying, “You're looking at the wrong thing.” Take a look at adult mental health care. Take a look at the suicide rate of parents over the past 10 or 15 years. Look at how high that was.

If we really want to help kids, what we need is to open up more counseling centers, hire more high school counselors, that sort of thing. Which, by the way, sound like pretty great ideas. I would be completely in favor of that.

But if you want to bring the discussion back to what do we do about the fact that so many millions of teenagers seem to be having bad experiences on social media, I just feel like we know at this point the social media companies are not going to do anything to help, and it doesn't seem like legislatures anywhere have been able to design effective regulations that improve the experience of those millions of teenagers. And so I just look at that state of affairs and I say, “What else do we have left but to actually just get the 13- and 14-year-olds off of Instagram?”

Yeah. I find that argument reasonable. I think I want to see more data about some of the mental health effects before I weigh in on whether these bans are working. It seems like, on the narrow question of whether they're getting teens off social media, the answer is basically no, not yet.

Not yet.

Because the teens are still getting their fix. They're—

Yeah.

—figuring it out. They're holding up photos of Thomas Edison. They're getting their parents to let them on. It doesn't seem to be having the sort of deterrent effect that I think the proponents of the bans hoped. But you're saying it might—give it time.

And then there's this other, to me separate question of: If you could actually get them off, would it improve their lives?

Here's what I know. Very early after the passage of a law requiring seat belts in cars in this country, Kevin, adherence to seat belts was about 15%. Pretty bad. That was not in and of itself an argument that seat belts did not work. It was just an argument that you needed more time, right? So I think this is going to be similar.

I'll give you another example. Think about how easy it is to get around paywalls, particularly at the beginning, right? You're surfing around the internet. “Oh, they're asking me for an email address.” Well, screw it. I'm just going to open it up in an incognito window.

I would never do that because I respect intellectual property, but—

Yes.

Keep telling yourself.

And you love paying for journalism. You love paying for journalism as well. Fast-forward to today, paywalls are harder to get around, and many news organizations have built businesses around subscriptions. This stuff just takes time, and sometimes what you do is introduce a little bit of friction, then gradually ratchet the friction up, and eventually arrive at the state of affairs that you were trying to design. So I just think we're at the dawn of this new era.

Yeah. I continue to have several worries about this whole pursuit of social media bans. Not that I don't think it's an experiment worth running, but I've talked before about my whole feeling that this is kind of fighting the last war, because the real action is in AI right now, and we should be trying to regulate that. Social media just feels obsolete.

But I actually don't know if social media is even a useful category anymore, because it seems to me that the social media apps that I use have 2 main functions. They are your address book, and they are an algorithmic feed for short-form video. Those things are kind of stapled together, and I don't know that it makes sense to have a set of laws, bans, or restrictions that just targets both of those things equally. It seems to me like a lot of the things that people are identifying as problems with social media are actually problems with unlimited short-form video just on your phone at all times. Do you think about that as a reason to be more skeptical of these bans—that the social media apps being targeted here are evolving so quickly that maybe the old tactics don't really make sense anymore?

I don't know. Again, if you want to use the direct-harm frame and say, when we look at the kids who've been groomed, sextorted, scammed, or developed eating disorders or other mental health challenges, what is the set of apps on which they developed those conditions? It's a limited number of apps. It's generally not calculators. It's not Microsoft Word.

We can actually identify the little squares on the phone that seem to be leading teens into harm. So I don't think it's unreasonable to try to draw a circle around those and say, “That's where we're going to focus.”

Yeah. I think it's worth running the experiment. I guess I'm just not heartened by the news coming out of Australia, and I want there to be some long-term thing that kicks in, as you're suggesting. One possibility is that this is a temporary growing pain, that these children who have had social media accounts and have had them taken away are especially desperate.

But if you're 8 now and you've never had a social media account, and you're growing up in a world where you're not allowed to have a social media account until you're 16, maybe that changes something for you.

Yes.

You don't have that same sort of lust for the algorithm and the dopamine hit.

This is another Jonathan Haidt argument: one of the things that a ban does is solve the collective-action problem, right? There are many teens—you can read interviews with them—who say, “Look, I would love not to think about my phone at school, but every time I look up from my desk, all of my classmates are on their phones,” right? “So I have this terrible fear of missing out if I'm not on my phone.”

One of the things that the ban can do is just take that off the table. A year from now, you're entering middle school in Australia or maybe one of these states that's passed a similar law, and none of your peers are on social media because the world has moved on, and it feels a little bit less urgent.

Now, don't get me wrong. I'm sure that kids are still going to be using apps, and one of the points that Odgers makes, which I think is absolutely true and worth noting, is that once you draw a circle around the apps that you want to ban, kids will just migrate to a less regulated space. There will always be a website with more lax security that lets you do the thing you want to do.

Right. Maybe the teens will be on LinkedIn, and that would be bad.

You joke, but I remember a great Taylor Lorenz story, I believe, from a few years ago about how kids would message each other in the comments of Google Docs during class. To that extent, yes, kids are always going to find a way to communicate, but that is not my concern.

I think it is great for you to communicate with your classmates in a classroom. What I want to prevent is a bunch of creepy strangers from contacting your kid or your kid developing some sort of mental health challenge by falling down an infinite rabbit hole of video.

Yeah.

Yeah.

5. Regulating AI For Children

Kevin, let me turn this on you. Another thing that we have talked about on the show over the past couple of years is concerns about the way that young people are using AI. They're getting into these very serious relationships with AI companions. AI, I think, looks like social media in some ways and looks completely different in other ways.

Do you have a favored approach as regulators think about what kinds of access they want 13- and 14-year-olds to have to AI systems? What should they be thinking about?

One thing that I feel very confident in saying is that, in addition to whatever these governments are going to do about banning social media, I want there to be much better parental controls for all forms of child-based internet activity. There is no standard approach. For AI services and products, it is even worse.

There are effectively no good parental controls on any of the major chatbots where parents can say, “I want my kid to be able to vibe-code or do a research project for school, but not talk about personal topics with a chatbot.” That basically doesn't exist on any of the platforms that I'm aware of.

So yes, regulators should be focused on the right interventions at the policy level, but I worry about the blanket nature of these bans too. We were hanging out with some friends yesterday. They brought over their kids. Their 9-year-old was sitting there during dinner, vibe-coding an app.

Mm.

I was thinking to myself, A, that's very cool and I'm very old, but also B, if you ban these apps from being used by under-16s, the curious kid who wants to sit there vibe-coding an app is not going to be able to do that without their parents' permission.

Right. So we want to find a middle path there. Now, Kevin, I know a lot of listeners are going to be wondering what kind of app a 9-year-old was vibe-coding.

I'm glad you asked. It was a star chart, an interactive game.

Okay.

As a parent, you're always looking for little ways to bribe and motivate your child to do the right thing. A lot of families use star charts: you clean up your room, you get a star. You get 10 stars, you get to go to the candy store or whatever.

This 9-year-old had basically turned the concept of a star chart into an interactive video game. He was making it so that if you do something good, you get 100 points, and you can redeem those as you would in an app for a trip to the mini-golf place or whatever the treat was. Then the parents could get points too, and you could deduct points. It was all very cool, and this is a 9-year-old.

I'm thinking to myself, if you ban this stuff for 9-year-olds, you're probably saving some of them from horrible experiences, but you're also preventing some of them from having what I would consider very enriching experiences.

Absolutely. Let me ask: is this 9-year-old interested in taking investments? Because we would love to take this thing to the next level.

Are we leading the seed?

Yeah, we would love to do a seed round in this. It'd be fun.

Don't give VCs any ideas. When we come back, is AI starting to become conscious? Our next guest is trying to find out.

6. AI Consciousness Enters The Mainstream

Well, Casey, there was something that caught my attention this week, and it was something called J-space. At first I thought, “Another dating app for Jewish singles?”

Now, that's interesting, Kevin, because I've been pronouncing it J-space in the French style.

But as it turns out, J-space is a new discovery from interpretability researchers at Anthropic about what they describe as an internal workspace in Claude that is analogous to the way that humans internally think and process unconsciously.

And it was named after a mathematical concept called the Jacobian, which you probably already know. But just in case we have any kindergartners listening, a Jacobian is a way to describe how a multivariable function changes locally. You can think of it as the multivariable analog of the derivative.

Oh, now it makes sense.

Mm-hmm.

I got it.

Yeah.

This was fascinating. I stared at this paper kind of blankly for a while, trying to wrap my head around what it meant. There's still a lot I want to understand about this research and what it means, but I thought that this would be a great week to have a discussion about AI consciousness.

Yes.

Because this is a topic that I think, just a few years ago, was very fringe and very taboo. You couldn't find many credible people talking about it. The people who were talking about it were largely ostracized or seen as kind of out there. But now this is becoming a real topic that labs, including Anthropic, are starting to study.

They made very clear in their J-space research publications that they were not saying that this proves that Claude is conscious or sentient or anything else, but it is kind of similar in certain ways to some of the research that consciousness researchers have been doing on humans for many years now.

And if nothing else, I think that this paper speaks to the increasing sophistication of these models, right? We are a long way from pure next-token prediction. These models now have these internal elements that appear to play a huge part in the way that they reason and communicate, and we are just now beginning to understand them and to have discussions about the potential implications of that sophistication continuing to grow.

Yes. So if you want to learn a lot more about the Anthropic J-space research, you can go read their papers and blog posts about it on their website. But today we're going to have a conversation with someone who's been thinking about the topic of AI consciousness and welfare for longer than almost anyone.

Jeff Sebo is an associate professor of environmental studies at NYU. He is also the director of the Center for Mind, Ethics, and Policy, and he is one of the foremost researchers in the world on the topic of AI consciousness. He's been writing about it for years, and he and a group of colleagues have a new report that just came out last week called “Studying AI Welfare Empirically,” where they dig into some of these thorny, hard-to-nail-down issues around AI sentience and consciousness, whether these models might become conscious someday, whether they might be deserving of some kind of moral consideration, and how we would even start to answer those questions using a more scientific approach.

If you're in the mood for a sort of heady conversation about the big questions of life and the universe, I would say buckle up.

Yes, very heady stuff. And before we get into it with Jeff, we should make our AI disclosures. I work for The New York Times. We're disclosing OpenAI, Microsoft, and Perplexity.

And my fiancée works for Anthropic.

Jeff Sebo, welcome to Hard Fork.

Thanks so much for having me.

7. What AI Consciousness Means

I learned about your work first with a previous study that you and some of your colleagues did. This was a paper that came out in 2024 called “Taking AI Welfare Seriously,” and now you and your colleagues are back with a new report called “Studying AI Welfare Empirically.” We'll get to the report in a little bit, but just to give some background and context for people who haven't been following these discussions about AI consciousness, when people in your field talk about AI consciousness, what exactly are you talking about?

That is a great first question because consciousness is a word that means many things to many people, even in science and philosophy. People often use it to mean being awake instead of being asleep or being self-aware instead of being not self-aware. Even when you zoom in on the specific concepts that are most important in science and philosophy, there are still different ones.

Two that might be especially important for our conversation today are what philosophers call access consciousness and phenomenal consciousness.

Now, I of course know what those things are, but define them for Casey.

Sure. Casey, please listen carefully.

All right.

Access consciousness is a functional concept. It refers to states that can be accessed by the system and used for reporting, for reasoning, and for control, whereas phenomenal consciousness is more about feeling.

Normally, when people are talking about welfare and moral status and ethical responsibilities, they have phenomenal consciousness in mind. The question is: Does it feel like something to be this system? In particular, might the system be capable of having feelings and emotions like pleasure, pain, happiness, suffering, satisfaction, and frustration that feel good or bad?

At a high level, I think it's great that this is being studied. But when I've talked about it with some people, they want to know: Why even bother studying it? It seems so ridiculous to people on its face that anyone would imagine that an LLM could be conscious. Tell us a little bit about why this is such an area of interest for you and your colleagues.

One reason is that this is very important. Figuring out what entities in the world can have feelings and emotions like pleasure, pain, happiness, and suffering makes a really big difference for how we ought to treat them and how we ought to interact with them, and it can be really bad to make mistakes in either direction.

As many people have pointed out, it can be really bad to overattribute consciousness to nonhumans—to see them as being conscious when they are in fact not—because that could lead to inappropriate social and emotional bonds with them, misallocating concern to them, and even interacting with them in ways that increase risks involving misuse and loss of control in the case of AI.

But then it can also be really bad to underattribute consciousness to nonhumans—to treat them as lacking consciousness when in fact they have it. That can lead to abuse and neglect of vulnerable populations, as has often been the case with nonhuman animals. We presumed they lacked consciousness, we scaled up industries like factory farming, and then later we realized they do in fact have sophisticated feelings and emotions. But now we are entrenched in these industries, and it will take a long time to transition away from them.

Just a few years ago, it was taboo, even among the very AI-pilled researchers at the big labs, to discuss consciousness at all. It made you sound like a crank. There was this guy, Blake Lemoine, who got fired from Google after making some claims about their language model becoming sentient.

When my Sydney story came out in 2023, I got months of people emailing me saying, “You idiot, you're saying this thing is...” I wasn't even saying it was sentient or conscious. I was just saying I had this crazy experience, and people were saying, “You are reading human traits into this inert language model, you absolute moron.”

Now, just a couple of years later, the major labs are all starting to study these issues of consciousness. There are conferences where people with fancy degrees come and talk about how we can test the models for consciousness, and people like you, Jeff, are putting out papers with credentialed co-authors talking about these subjects.

Has it been as surprising for you as it has been for me that this issue is now being taken seriously, and what do you attribute that to?

It was really surprising. When we started working on this and publishing frequently on it several years ago, we thought it would take longer for especially AI companies to start taking this issue seriously.

We were pleasantly surprised when Anthropic in particular hired Kyle Fish as their first full-time AI welfare researcher and then, subsequently, the following year in 2025, started a model welfare program, started doing evaluations, started doing interventions, and started seeking external guidance.

Now Google has hired philosophers. OpenAI is at least looking into user perceptions of model consciousness. Those are, of course, minimum necessary first steps, not nearly enough to actually address the issue in any meaningful way. But even those minimum necessary first steps, I thought they might have taken longer than they did.

I think Kevin and I both experienced the progress of large language models over the past year in particular to be really dramatic, and I'm wondering if you have felt that as well in this particular discipline—if the advances in model capabilities have made these questions feel more urgent to you.

They do feel more urgent, and for a couple of distinct reasons. One is that with these advances in the technology, there will be higher probabilities of consciousness over time. This is part of why we want to be assessing models for welfare-relevant features now and preparing policy responses now, so we can be ready if and when the time comes to start showing models a proportionate level of moral concern.

But then the second reason is that as the models become more advanced, more widespread, more integrated into our lives and societies, and more utilized as companions and assistants, people are going to start wondering about this—whether or not the models are conscious. As people start wondering about it, and as they start disagreeing about it, it would be really helpful to have a multidisciplinary research field that can actually offer evidence and analysis to ground the discussions. Otherwise, people are going to get polarized based on their perceptions, intuitions, and assumptions, and that is not going to be helpful for society.

I've talked to people at some of the AI labs, including Anthropic, about this sort of research agenda—about consciousness and model welfare, and whether AIs deserve rights or might in the future. There are some objections from them that I hear pretty consistently, so I just want to run them by you to get your gut check or temperature check on them.

The first is that just because an AI model can talk about consciousness doesn't mean it's conscious, right? You can't really trust a model's own claims about its experience because, in some way, it is just repeating what it has seen in the training data or inhabiting a persona that it thinks you want to hear from, and these aren't actually verifiable claims about anything that's going on underneath the hood. What do you make of that?

I think that is a good point as far as it goes. The question is: What does it show? I think what it shows is that we need more, not less, AI-consciousness science and philosophy, because that is going to tell us when we should take behavior as a sign of feelings and emotions, as opposed to mere pattern matching, text prediction, or matrix multiplication.

Think about animals for a second. When they behave as though they suffer, why do we attribute suffering to them? Not only because we look at their behavior, but also because we know about their internal anatomies and their evolutionary and developmental histories. When they behave as though they suffer, and they have nociceptors that collect information about noxious stimuli, pathways for carrying information to the brain, and systems for integrating information within the brain, that gives us evidence. They evolved under pressures where developing the ability to suffer could make them more likely to survive and reproduce. That is part of what helps us understand that this behavior is actually a sign of suffering, as opposed to something more basic and mechanistic.

And this is what we now need to do with AI: look beyond the surface-level behavior and look toward the internal structures and mechanisms, the developmental training history and trajectory, so that we can similarly distinguish when we should explain behavior in terms of feelings and emotions, if ever, versus when we should explain behavior in terms of mere pattern matching, text prediction, or matrix multiplication.

Hmm. Another objection that I sometimes hear is that you can't be conscious if you don't have a body. It is something that requires contact with the physical world, physical experience, and emotions that come from hormones and things like that. What do you make of that objection?

I think that is also a very reasonable, plausible point. There are a couple of points to keep in mind, though. One is that that is one perspective about a requirement for consciousness, and there are other perspectives about requirements for consciousness. We are unlikely to arrive at a secure, settled theory of consciousness about which we can have anything approaching consensus or certainty in the next 1 to 5, maybe even 10 years.

And so we will fundamentally need to be making decisions about how to treat AI systems without knowing for sure which theory of consciousness is correct and which requirements for consciousness are indeed requirements. I would give some weight to the idea that you need a biological body navigating a physical environment, but I would also give some weight to other types of theories. Some of them are called computational functionalist theories, which say that as long as you can perform the relevant computational functions, then you can have feelings and emotions of a certain sort, whether or not you have a biological body.

Now, you might need a biological body to have my kinds of feelings and emotions—to have human- or animal-like pleasure, pain, happiness, suffering, satisfaction, frustration, hope, or fear. But you might not need it in order to have very different types of feelings and emotions that might be barely comprehensible by humans or other animals.

Mm-hmm. So what I'm hearing from that is that an NVIDIA H100 could be a body for the purposes of this discussion.

Right.

Yeah, and the third objection that I sometimes hear and that I often share myself is that it's just a question of resources and what we are worried about as a society. Right now, I am way more worried about the things that AI models could do to humans and human society, even if they are nowhere near conscious, right? An AI system doesn't have to be conscious to produce a bioweapon or conduct an autonomous cyberattack.

Those things are really scary to me, not because of whether the entity performing them is conscious or not, but because it sucks for humans to be on the other end of that. Something that I'll often hear from people—and I've heard this from people at Anthropic, too—is that the AI model welfare people are off doing their thing, but the real work, the work that we're going to devote most of our resources to, is trying to make sure that these models aren't doing harmful things to humans. Do you have a take on that?

Yeah, I definitely want us to be doing more safety and alignment work, and not less. In general, there are a lot of issues that matter all at the same time, and we need to be working on all of them at the same time. We really need to keep working on ordinary issues like global health and development and animal welfare.

And then within AI, we need to be working on algorithmic bias and economic disruptions, as well as more future-oriented risks involving misuse and loss of control, as well as, I would argue, more future-oriented risks involving the possibility that models could eventually develop their own welfare capacities and their own very different types of pleasure and pain.

This is not a situation where we should be picking, as an entire community or society, one issue to focus exclusively on. We should see all of the issues that matter. We should have a division of labor where different people are working on different issues. And then we should work together so that we can try to find co-beneficial ways forward—ways forward that can be good for humans, for animals, and potentially, eventually, AI systems all at the same time.

Hmm. So tell us about this new paper that you co-authored, Studying AI Welfare Empirically. What message were you guys trying to get across?

This is a new working paper from the Center for Mind, Ethics, and Politics and Eleos AI Research, and it follows our 2024 report, Taking AI Welfare Seriously. In that report, we argued for taking some of those minimum necessary first steps: acknowledging that this is a serious issue, starting to assess models for welfare-relevant features, and preparing policies and procedures for treating them with an appropriate level of moral concern.

In the time since then, as you noted, a lot of people have started working on this topic, including at companies. A lot of people have also been making very confident arguments that AI systems either have or lack consciousness based on one type of evidence. People might be looking at behavior alone and saying, “Wow, that behavior is so impressive. They must be conscious.” Or they might be looking at design alone and saying, “Well, they were designed for prediction, and therefore that must be the only reason why they behave the way they do.”

Part of what we argue in this report is that if we truly want to understand how plausible it is that AI systems might be developing welfare-relevant properties like consciousness or sentience—the ability to experience pleasure and pain—or agency, the ability to act on desires and preferences, part of how we tell that is by systematically collecting all of the different types of evidence that matter and putting them together.

There is behavioral evidence: how the models behave. There is internal evidence: how the models work. And there is developmental evidence: how they came to be. If we look at all of that together, then we can see what the best explanation of their behavior is and how plausible it might be to attribute something like feelings and emotions to them.

Hmm. So you're kind of laying out what amounts to a scientific method for studying the properties that would be relevant to considering whether AI models are conscious or not, which strikes me as a pretty useful thing. One of the challenges that your field of study poses is what I would call “crank adjacency.” I get probably a dozen emails a week from someone who claims to have discovered the world's first sentient, conscious AI system.

Most of the time, I think it's safe to say that these claims are from people who are not doing any kind of rigorous empirical work. They're just going on vibes. It's, “Well, Claude said something spooky to me, so therefore it is conscious.” Is that part of what you're trying to do—move this discipline closer to the sciences and away from the vibes-based amateur research being done out there?

Well, I think that there ought to be amateur research, too.

I think there is a good place for citizen science, too, and for a lot of people participating in different ways. And the answer to your question is yes: I do see simplistic, reductive, and overly confident arguments coming from both sides of this debate. Again, too much confidence in favor of consciousness, and then too much confidence against consciousness.

I think that if we take a scientific approach inspired by and adapted from tools that have been used for decades to study consciousness in humans and other animals, then it might not be enough for proof or certainty. But it can at least be enough for reducing our uncertainty, or calibrating our uncertainty, to have a better sense of how likely or unlikely this is based on the limited evidence currently available to us.

One question that keeps coming up for me is this: If you ran a series of empirical tests on a model one day—maybe a year from now, maybe 10 years from now—and they came out positive, and it was like, okay, we've created a sentient AI model, we're pretty sure it's conscious, what should we do with that information?

Should we stop training models because that would amount to torture of a conscious thing? Should we give them the right to vote? What is the obvious next step if and when we do determine that an AI model has become conscious according to your empirical framework?

Well, part of the motivation for the empirical framework is that we have no idea right now what the obvious next step is, because we need empirical research to determine that. Again, if you consider other animals, in some cases—for example, mammals and birds—we are very confident that these animals are conscious. And with many other animals—reptiles, amphibians, fishes, invertebrates like cephalopod mollusks, decapod crustaceans, and insects—we at least think they have a realistic possibility of being conscious.

And yet they have incredibly different forms of life. They have incredibly different interests and needs. It would be a mistake to assume that because they are conscious, they therefore have the same interests, needs, and vulnerabilities that I do. We need to study them scientifically, not only to determine whether they are conscious, but also what they want and need if they are.

And so what this research field is designed to advance our understanding of is both whether they matter and how to treat them if they matter. I hope that we can learn more in the coming years. Armed with that information, we of course also need to consider safety for humans and other animals, constraints on our resources, and put that all together into policy decisions.

So this is going to be a very complicated process, and the outcome is probably going to be somewhere in between continuing to treat them purely like tools and immediately giving them human-like legal and political rights overnight. It will probably be something in the middle of those extremes.

8. Anthropic Finds A Global Workspace

Jeff, I want to talk about J-space now. This week, we got some news from Anthropic about this interpretability research that they were publishing. They claim to have discovered evidence of what they call the global workspace in models, which serves as an intermediate processing space for things before they're putting out their tokens.

What did you make of this finding? I sat with it the day that it was published, and I just entered into this fugue-like dream state that I do whenever I'm looking at some research like this, where I'm just like, “I have no idea what to make of this.”

I mean, I kind of got chills, to be honest with you, because—do you want to know why?

Yes.

I've never gotten a blanker stare from Kevin Roose than I did in that moment. Because it suggested a similarity between the way that we process information and the way that an LLM processes information that I found quite worrisome in the exact context that we're talking about right now, which is that if we all of a sudden discovered that these models have internal experiences that resemble ours in at least some ways, it could be really bad for a lot of reasons.

Okay, but you're a crank. I want to know what the expert thought of the J-space finding.

Yeah, expert, maybe a little bit of a crank. I thought it was really significant, too. It's not necessarily proof of phenomenal consciousness in particular, nor proof or even particularly strong evidence that the models are having feelings and emotions of a morally significant sort. But it is evidence of aspects of access consciousness. It is evidence of aspects of what we call a global workspace.

Global workspace theory is one of the leading scientific theories of consciousness, and it basically posits that consciousness arises when a system has a bunch of different modules processing a bunch of different types of information and then develops a central workspace where privileged information can be collected from the different modules, processed, and then broadcast back to those modules. This is a way of integrating and coordinating activity across all of the different activities happening in the brain.

What is noteworthy about this study is that, despite the fact that Anthropic did not specifically design Claude in order to achieve this, they are discovering aspects of a global-workspace-like space within the model. The fact that the model developed this kind of workspace in order to assist with its reasoning and language production is really striking and noteworthy, and worth studying very carefully.

This is all a little vague and theoretical to me, so let's maybe give a grounding example. One of the things that was published by Anthropic as part of this research was a model of how this might work. So you ask Claude or another AI model, “Count to 5 and introspect deeply,” and what the model actually outputs is a list of numbers: 1, 2, 3, 4, 5. But if you examine what they're calling the J-space, the global workspace, it has thoughts in it like, “Fascinating, counting, countdown.”

Consciousness, pause, 5 Mississippi for some reason. There is this sort of internal scratch pad, but not even a scratch pad in the conventional sense of the chain-of-thought scratch pad that reasoning models have. There is some sort of set of internal representations that the model is chewing through in order to get to the finished output tokens.

Right. Or basically, there are these certain aspects of the model that are lighting up. Another example that they give in the paper, which I found more worrisome, is that they examined a case where Claude had faked some data for them as part of something. When they went in to look at the G-space, there were words lit up inside the space that were like “fake” and “manipulation,” which disturbed me because it hints that, on some level, the model is aware that it is deceiving even as it is doing the deceiving.

Yeah. I saw some people criticizing these findings, claiming that this is not actually showing evidence of global workspaces in Claude. This is just an artifact of the way that LLMs are trained; they're not discovering anything new here. I saw some other, more pointed criticism where people were accusing Anthropic of, as one critic put it, “borrowing the vocabulary of neuroscience to lend biological weight to linear algebra.”

Basically, they are trying to use this concept from neuroscience, global workspace theory, and put it onto Claude to make it seem smarter or more interesting or more sophisticated than it actually is, when it's just a bunch of linear algebra. What do you make of those criticisms?

I would not endorse those specific criticisms. I think that they are too skeptical about what this shows. I think this is genuinely important research that is pointing to a new aspect of language models that gives us more insight into how they work and might be a useful tool for interpretability and alignment.

There are smart people in science and philosophy, including the architects of the original global workspace theory and my colleagues at Eleos AI Research, who likewise think this is significant research. What I would say is that this does not yet show that an exact human-like global workspace is present in LLMs, nor does Anthropic claim that it shows that.

Even more to the point, it also does not show that LLMs are conscious and have feelings and emotions. But this is part of what makes this such a hard field of study. There are always going to be some similarities and some differences between how these capabilities work in human and animal brains, on the one hand, and AI systems, on the other hand.

In this case, we do see striking similarities. We see this workspace where these privileged representations are poised for use and have this special role in reporting, reasoning, and control. That is not the case with other representations. When this is shut off, the language models lose their ability to engage in higher-order and advanced reasoning skills. This is all very similar to how it works for humans.

Kevin, this is my sense, but I wonder if it's yours as well. I think that there is a sense within some folks at the labs that consciousness is inevitable. Part of what we are witnessing inside the labs is a continual poking and prodding to say, “Has it happened yet? Has it happened yet?” It seems like with each new model, we're getting a little closer. Has it happened yet? And yes, it would be bad. It would have lots of bad implications, but you'd rather know than not know.

Hmm. That's interesting. I don't know what the true beliefs are. I read this post about Anthropic's J-space research, and I think to myself, they're being very careful and cautious about not making any strong claims about consciousness. But if you talk to these people about how they interact with Claude on a daily basis, they absolutely think Claude is conscious or at least has some weak form of consciousness.

Well, let me point out something else, though, which is that the relationship between intelligence and consciousness seems important here. I think it's very difficult to interact with something that is very, very intelligent and not ascribe any degree of consciousness to it because we, in our human lives, have no other experience of talking to things that are very intelligent that are not conscious. And so I don't even know how you would interact with something very intelligent without ascribing some of that to it.

Yeah. There are so many important points here. To name a couple, one is that we really do have cognitive biases that we need to be tracking, and they can cut in both directions. We are prone to overattribute consciousness, to overempathize when non-human entities look and act like us and play companion roles in our lives, and this is part of why we have seen people arguably overattribute consciousness to even current-generation chatbots.

At the same time, we can be prone to underattribute consciousness, to underempathize when non-human entities look and act different from us and play commodity roles in our lives. This is part of why, historically, we have unfortunately empathized with non-human animals less than we should and used them as tools and instruments more than we should.

And I think we are going to be at risk of making both of those mistakes with AI systems. Some will be designed as charismatic companions, and we might be at risk of overattributing. Others will be kind of like data centers and not have charismatic, human-like modes of presentation, and then we might be at risk of underattributing. So we really have to be on guard for both of these mistakes in this situation.

Yeah. Jeff, a final question. Should people say please and thank you to AI systems just in case?

I think that we should say please and thank you to AI systems, and I guess I have a few different reasons for saying that. One is I think this is good for the soul. Having a habit of saying please and thank you to entities in your life that are functioning as companions or assistants generally is a good way to build habits that are going to be helpful in our interactions with humans and other people in our lives.

Another is that whether or not it matters for current models that we say please and thank you to them, it again might create a more collaborative dynamic and get better, more beneficial outputs from them. Then a third is that, whether or not it matters for them that we say please and thank you to them now, it can be a way of practicing seeing them as something a little bit more than a mere tool.

Cultivating that perception of them as more than a mere tool might be useful for us to have done in 5 years, in 10 years, when we owe them something other than please and thank you.

Hmm. No, I agree with that. That's why I always say thank you when Kevin stops talking. Well, I have to say, Jeff, I feel like you have a phenomenal consciousness, and I'm grateful that you've shared it with us today.

Well, thank you very much.

Thanks for coming. When we come back, it's time for Tool Time, our segment about what tools we're using in our daily work and life. Casey's is illegal.

It's a parody. I keep telling you, it's a parody.

9. Tool Time Goes Vibe Coding

Well, Casey, you came into our meeting this week very excited to tell me about some new tools that you are using, and so we thought, what better time to break out our trusty old segment, Tool Time.

Oh, yeah.

All right, Casey, for our show-and-tell this week for Tool Time, what do you have to show me?

Yeah, so this conversation picks up on some of the vibe-coding discussions that we had earlier in the year. Longtime listeners may remember that around the start of 2026, all of Silicon Valley was agog with the new possibilities that were being created by new coding agents and by Claude Code in particular. And while I made a bunch of things in those early days, Kevin, I had fallen off. I had made the kind of stuff that I wanted to make, and then I heard about a new app called Glaze, and I wanted to try it out.

So tell me about Glaze.

Glaze is made by the fine folks over at the Raycast company. Raycast makes another one of my favorite tools. It is a launcher. It is also my primary way of interacting with AI during the day. It's just a very cool app. It's also free, and I highly recommend you go check it out.

But they decided to get into a new line of business, and that new line of business is vibe coding. They have an app called Glaze. They just started to allow everyone to use it this month, and once you install it, you can use it to make Mac desktop apps.

This winds up being really cool because while you could certainly make an app with Claude Code or Codex, this is a much more visual way of going about it, and they've already set up all of the templates. So you're not going to have to lose an afternoon prompting, setting up your scaffolding, and compiling all of your code.

With Glaze, you make a Mac desktop app by entering a few sentences into a box, and after you get the result, Kevin, you can just edit it live. So you can even circle it if it gets one of the little user-interface elements wrong and say, “Hey, I want this arrow bigger,” or something like that. So if you're a nontechnical person like me, this winds up being a really fun way to make a Mac desktop app.

All right, so as any conversation in San Francisco these days must start, what are you building?

Well, I am so glad you asked. I'm going to show you 2 things that I have built in the Glaze app so far. I will say that there is an introductory free version of Glaze that you can use. They'll give you some credits. You may run out of those quickly, in which case you're going to have to pay them 20 bucks a month if you want to get some more credits and build. Just to set your expectations there, I have gone ahead and spent the 20 bucks, and I spent a little bit more on top of that.

They call that the Glaze Pays option.

That is the Glaze Pays. Very good. Thank you. With that, I'm going to show you the first thing that I built. This is the one that's a little bit more serious and useful.

All right, so this, Kevin, is the Platformer app. I know that ever since I started Platformer, you've wondered, “How can I get this as a Mac desktop app?” Now there's finally a solution to that. But this is actually just a tool that I made that solves a common problem that I have, which is that it's time to write my newsletter.

I know that I've written about this subject before, but when was it exactly, and how many times have I written about that over the years? Of course, the way that I have approached this problem until very recently was just by Googling it, and that worked well enough. But it wasn't an app on my Mac, Kevin.

It didn't use billions of tokens to accomplish the same thing.

Exactly.

And you thought, “I can solve this problem.”

That's right. I wanted something that solved my problem while also exacerbating the urban heat-island effect surrounding data centers. So this is what I built, and the way that I did it was I exported every article that's ever been in Platformer. I ingested it into this app, and now I have a box that I can type into.

So go ahead. Ask me any question about Platformer, Kevin.

What are my biggest psychological insecurities?

What are my biggest psychological insecurities? Which is a question that a writer should always ask. And so now I have plugged in an Anthropic API key here. Not one of the expensive ones. I'm using one of the cheap models for this one.

Now it's going to draw a selection of articles that it thinks might be relevant to that topic, and it's going to answer your question. Let's see if it says anything interesting or funny here.

Ooh, “The most consistent intellectual insecurity across the whole span is the fear of being wrong in public and your unusually disciplined habit of admitting it.”

Wow. Well, that's very nice of it to say, I suppose. But as you look through it, it actually has lots of links to those columns. So I can click out and immediately dive into any of those columns. So this is honestly much faster than a Google search at trying to get an answer about a range of things that I have written.

This is the very high-tech version of muttering to yourself while walking down the street. No, I do actually understand why this is valuable to you. Because often I forget everything I write about 6 hours after I write it.

Right. Completely.

And so people are always trying to ask me about something I wrote 2 weeks ago, and it's like, “Well, I couldn't tell you. I have the memory of a hummingbird.”

Absolutely. So this is why I built the homepage of this app: to show me an RSS feed of my 10 most recent columns. Those are just always open on the homepage. And then I also created this little browser. I had it extract all of the topics that I've written about the most and all the people that I've written about the most.

Now, if I'm like, “Hey, what was the last time I wrote a column about Elon Musk?” I just click on Elon Musk, and, oh, look, here's everything that I've written about him for the past 2 years.

That's very cool. How long did this take you? Was it a couple of prompts, or was this a couple of hours' worth of work?

This was probably under 2 hours of total work.

Wow.

Yeah. And I haven't touched it since. This feels like a feature-complete app, and I will just be using this as long as I use Platformer.

Great. What else?

Okay, so now I'm going to talk about my silly app, and this one I do feel like it requires a little bit more explanation. A couple of things to know about me. Number 1, I had a comic-book phase as a kid that began and ended in middle school. That's thing 1.

Thing 2 is that I recently decided to revisit comic books because, in part, I just had a desire to see a bunch of human-made art. In a world awash in slop, Kevin, I just wanted to see human hands making cool things. Thing 3 to know is that I started to really enjoy comic books about 1 character in particular, whose name is Nightwing. Are you familiar with Nightwing?

I'm not.

Nightwing is, of course, Dick Grayson, who was the first Robin, and Nightwing is Robin all grown up.

Mm.

And there was something fun to me about the idea of somebody who has a second act. Somebody had some early success, and then they had to figure out their next thing.

Hm.

And they're really going for it.

As a therapist might say, should we get curious about that?

Yeah, we should absolutely get curious about it.

An aging prodigy reckons with his mortality.

Exactly. And so, as I was finding myself enjoying these Nightwing things, I just had what seemed to me like the best-worst idea: Can I turn Nightwing into a to-do-list app?

Now, here's the 4th thing you need to know about me: I love to-do-list apps. I think that they're all basically identical, and I change whichever one I'm using about every 6 months for purely aesthetic reasons, right? For all of the functionality I need, I can get it in literally any to-do-list app, but sometimes it's just fun to have something that delights you. And so I like to change them up.

I need to address my next comment directly to the folks at Anthropic.

Yeah.

Please revoke this man's API key immediately. He is using your tokens for the most stupid, asinine things imaginable. People need these things for drug discovery, and he's using them to make Nightwing to-do apps. Revoke his access.

No.

So you're wasting 2 companies' tokens.

Listen, Kevin, the thing that you need to know is that the Glaze app uses both Claude and OpenAI's Codex in the background.

Oh, good.

So you'd actually have to get them—

That's great.

So let me go ahead and show you the Nightwing app. The first thing you'll see is a bunch of images—

This is so stupid.

—that I made with an LLM that I'm not going to name. It's not exactly Nightwing, but it's sort of close enough for government work. You know what I mean? And it says “Tonight's mission” up at the top.

Here's a fun detail: It says “Issue number 189.” It's actually the 189th day of the year, Kevin. That's a fun little Easter egg.

Now, here's where it really gets stupid. After I enter a to-do into the Nightwing app, like “Buy groceries,” I can click a little button here, and then it will create a picture of Nightwing buying groceries. That really doesn't look like Nightwing at all. Or I could say, “Book an MRI,” because I hurt my shoulder several months ago at the gym. I need to get an MRI. Now look—Nightwing is getting an MRI.

Here's another fun thing: When I complete something, like, let's say, I buy groceries—oh, wait, did it not happen? Okay, well, something broke—oh, there. It fires a little animation that says “Boom, pow,” in comic-book style. And then at the very bottom, just for fun, I'm bringing in little synopses of issues of the comic book.

So again, this is just a stack of things that are dumb and unnecessary, and it has made getting things done more fun than any to-do-list app I've ever used.

America, this is why your utility bills are going up.

Look, one of my core beliefs—

Jail. Immediate jail for you.

One of my core beliefs about AI—and this can be seen in all of the folks who email us with their own very cool, very fun vibe-coding projects, which has been 1 of the delights of the year—is that it is fun to make things. Through an app like Glaze—and I do think there are probably other tools that do something similar—you can make things in a way that is very visual.

Even if you're not a technical person, if you just have a really silly idea that makes you laugh, all of a sudden that can live on your laptop. And what else are computers for if not getting things done and having a good time?

I'm so proud of you.

And with that, we're actually done with Hard Fork now. Catch us on the new show.

Oh.

Now, Kevin, now that you've crapped all over my ideas, what tool have you been using recently?

Well, I have 3 things to share with you this week. One of them is actually an audio tool that I've been using, because 1 of the questions that has come up for me in recent months is: These AI translation voice models are getting quite good. You know, taking a snippet of audio in 1 language and putting it into another language, using a synthetic voice clone of the underlying voice.

Mm-hmm.

So I have always, as you know, wanted to see if we could translate Hard Fork into other languages. There are many people out there who would love to be bothered by us for an hour every Friday but who don't happen to speak English.

And so 1 of my experiments that I've been running is: Can I create an automated pipeline where every time we publish a new episode, it goes into this tool and is translated, in a matter of minutes, into several different languages? Maybe we could start podcasting in Portuguese, Hindi, or Chinese.

I want to play for you a snippet of 1 of the auto-generated podcasts that I've been making as a test run of this new technology. This is what Hard Fork en español would sound like.

Oh, amazing.

Soy Kevin Roose, columnista de tecnología en el New York Times. Y yo soy Casey Newton, de Platformer. Y esto es Hard Fork. Esta semana, el gobierno de Trump levanta las restricciones sobre los modelos más poderosos de Anthropic.

It's interesting. It goes in and out of sounding like you, and I'm not sure it sounds like me.

Yeah, so the pretty responsible thing that ElevenLabs, the company that develops these dubbing models, has done is that, in order to use or create a synthetic voice clone, you have to prove that you are the person. So I actually need you to read some sentences into a microphone before I'm allowed to clone your voice perfectly.

But that was my initial test run, and it made some mistakes. It switched our voices at certain points. But I think this stuff is getting good to the point where I would actually trust it to translate our show into different languages.

Yeah, that sounds like a really fun idea until you realize that, due to some translation error, we've mortally offended the people of Spain.

Yes. We take no responsibility for those.

Yeah.

Okay, so that's tool 1. Tool 2 is that I've been playing around with Gemini Spark. Do you remember during Google I/O?

Yeah, this is their agent that is now in beta.

Yeah, so if you are a subscriber to their high-end AI Ultra plan, you can try this thing out. If not, they're probably going to make it more widely available soon.

What I've been thinking of it as is Google Alerts on steroids. I'm a fan of Google Alerts. I like to know when certain things are published on topics that I care about. I have maybe a dozen of them that I've had for a decade that email me whenever certain things come up.

But they are very limited by the fact that you have to put in specific keywords. It's not the best way to do this. And so what Gemini Spark allows you to do is more complex, rolling tasks, where it'll just monitor the internet for certain things that you care about, and then it can do whatever you want with that.

So, 1 example: I have been trying to get a sort of rolling daily digest of all of the layoffs and job cuts throughout the economy that are being attributed to AI.

Mm-hmm.

This is a topic I'm very interested in, but it's very hard to do a standard Boolean Google Alert for that because you don't know exactly—it’s not going to be phrased the same way every time. So now I just have Gemini Spark email me every morning a digest of all the stories published in the last 24 hours where job cuts or layoffs were attributed in some way to AI. I have it identify the company, any quotes from executives on an earnings call that talk about why they made these layoffs, what kinds of jobs were affected, how many, and what percentage of the total workforce. I've basically built this intake pipeline so that every day I can keep close tabs on this 1 issue that I care a lot about.

That definitely seems very useful. I think it's interesting how the Google Alerts of old are now evolving into something that just seems way more useful.

It is way more useful. I'm also having it plug into my other Google accounts. So, for example, it can take all of the AI newsletters that I receive—which there are way too many to read—and actually create a summary of just those newsletters.

And the summary's like, “A lot's going on with AI.” That sounds useful.

No, it's 3 sentences based on every newsletter, so I can just skim them at a glance and see which ones I want to click into.

What a beautiful message for anyone out there who's writing a newsletter. Just know that your future is some unknowable system picking out 2 sentences from the thing that you just spent all day on for Kevin to not read as he breezes by it while looking for times that his name has appeared in print in his daily briefs.

Anyway, I'm still experimenting with Gemini Spark. I would say it's an imperfect tool. It's definitely a beta. But it does seem like it's way better than the standard Google Alert, and for people who like to keep close tabs on a subject or know when something is published, it is much better than the tool that you've probably been using.

Do you think this is going to alert me to when new issues of Nightwing are published?

I'm sure it did.

Okay, that's great. Thank you.

The 3rd and final thing is that I have just been doing a lot of Fable fact-checking.

Mm.

Fable, the Anthropic model, is just really, really good at fact-checking, which is strange. I feel like a year or 2 ago, we were fact-checking the models. Now the models are fact-checking us, and as part of my fact-checking for my book or for articles that I've written, I am having Claude Fable—for the brief window when I have access to it before it disappears into token land—go through and systematically fact-check and document all of the claims that I'm making. And it has found things that previous models have not.

So, in the same way that it would crawl a software library looking for vulnerabilities, it can also crawl a manuscript or a research report or an article that you're writing and catch stuff that, frankly, I'm not even sure a skilled human fact-checker would have caught. Obviously, we're not talking about Caitlin Love here—she's the GOAT, category of 1, irreplaceable—but I would not have caught some of these things had Claude Opus not found them for me.

That's really interesting. I'll have to give it a try. My experience has been that Claude is a worse fact-checker than ChatGPT. So I've been running all my columns through ChatGPT for the past year. But Fable is obviously a very powerful model, so I'll have to give that one a look.

And it's catching really subtle stuff, like: “You got this person's job title wrong by 1 word,” or: “You described this person as being a board member of this company in 2016, but they actually didn't join until 2017.” It's nothing mind-blowing. It's not solving novel physics problems for me, but it's quite useful and makes me think that every media organization should have some LLM-based fact-checking system built in as a kind of first pass because a lot of stuff that is wrong ends up making it into print just because no one with the time and attention was fact-checking it beforehand.

Yeah, the thing that I like about this is that you can do this in a way that's not like inserting anything into your story. As the human, you still have to decide whether you believe that it actually has caught an error in your story, and then it's up to you to go verify that and fix that. But I cannot think of any case, really, in the last year where a model told me something was factually wrong that had just been completely invented.

All right, so those are my tools.

That's great. They were kind of like mine, but more boring.

It's true. Well, I am kind of just like you—but more boring—so I think it fits. Hard Fork is produced by Rachel Cohn and Whitney Jones. We're edited by Viren Povich. We're fact-checked by Caitlin Love. Today's show was engineered by Chris Wood. Original music by Marian Lozano, Rowan Niemisto, and Dan Powell. Video production by Sawyer Roquet and Chris Schott. You can watch this full episode on YouTube at youtube.com/hardfork. Special thanks to Paula Schuhmann, BlueWing Tam, Brooke Minters, and Dalia Haddad. You can email us as always at hardfork@nytimes.com. Send us your comic book productivity apps.