Hard Fork - DeepSeek DeepDive + Hands-On With Operator + Hot Mess Express!

Episode Date: January 31, 2025

This week we go even deeper on DeepSeek. ChinaTalks’ Jordan Schneider joins us to explain the Chinese A.I. industry and to break down the reaction inside of China to DeepSeek’s sudden success. Then … hello, Operator! We put OpenAI’s new agent software to the test. And finally, all aboard for another ride on the Hot Mess Express! Guest:Jordan Schneider, founder and editor in chief of ChinaTalk Additional Reading:Why DeepSeek Could Change What Silicon Valley Believes About A.I.OpenAI launches its agentFable, a Book App, Makes Changes After Offensive A.I. MessagesAmazon Pauses Drone Deliveries After Aircraft Crashed in Rain  Unlock full access to New York Times podcasts and explore everything from politics to pop culture. Subscribe today at nytimes.com/podcasts or on Apple Podcasts and Spotify.

I just got my weekly, you know, I set up ChatGPT to email me a weekly affirmation before we start taping, because you can do that now with the tasks feature. Yeah, people say this is the most expensive way to email yourself a reminder. So what sort of affirmation did we get? Today it said, you are an incredible podcast host, sharp, engaging, and completely in command of the mic.

Your taping today is gonna be phenomenal, and you're going to absolutely kill it. Wow, and that's why it's so important that ChatGPT can actually listen to podcasts because I don't think it would say that if it had actually ever heard us. It would say, just get this over with. Get on with it.

I'm Kevin Ruse, a tech columnist at the New York Times. I'm Casey Noon from Platformer. And this is Hard Fork. This week, we go deeper on DeepSeek. China Talk's Jordan Schneider joins us to break down the race to build powerful AI. Then, hello, operator, Kevin and I put OpenAI's new agent software to the test. And finally, the train is coming back to the station for a round of Hot Mess Express. Well, Casey, it is rare that we spend two consecutive episodes of this show

talking about the same company, but I think it is fair to say that what is happening with DeepSeek has only gotten more interesting and more confusing. Yeah, that's right. It's hard to remember a story in recent months, Kevin, that has generated quite as much interest as what is going on with DeepSeek. Now, DeepSeek, for anyone catching up, is this relatively new Chinese AI startup that released some very impressive and cheap AI models this month that lots of Americans have started downloading and using.

Yeah, so some people are calling this a Sputnik moment for the AI industry when kind of every nation perks up and starts, you know, paying attention at the same time to the AI arms race. Some people are saying this is the biggest thing to happen in AI since the release of chat gbt but Casey why don't you just catch us up on what has been happening since we recorded our emergency podcast episode just two days ago well i would say that there have probably been three stories kevin that i would share to give you a quick flavor of what's been going on one a market research firm says deep seek was downloaded 1.9 million times on iOS in recent days,

and about 1.2 million times on the Google Play Store. The second thing I would point out is that DeepSeq has been banned by the US Navy over security concerns, which I think is unfortunate, because what is a submarine doing if not DeepSeq-ing? It was also banned in Italy, by the way, after the data protection regulator made an inquiry. And finally, Kevin, OpenAI says that there is evidence that DeepSeek distilled its models. Distillation is kind of the AI lingo or euphemism

for they used our API to try to unravel everything we were doing and use our data in ways that we don't approve of. And now Microsoft and OpenAI are now jointly investigating whether DeepSeek abused their API. And of course we can only imagine how OpenAI is feeling about the fact that their data might have been used without payment or consent.

Yeah, it must be really hard to think that someone might be out there trading AI models on your data without permission. And I want to acknowledge that literally every single user of Blue Sky already made this joke but they were all funny and I'm so happy to repeat it here on Hard Fork this week. Now, Kevin, as always, when we talk about AI, we have certain disclosures to make. The New York Times company is currently suing OpenAI and Microsoft over copyright violations alleged related to the use of their copyrighted data to train AI models.

I think that was good. That was very good. And I'm in love with a man who works at Anthropic. Now, with that said, Kevin, we have even further we want to go into the DeepSeek story and we want to do it with the help of Jordan Schneider. Yes, we are bringing in the big guns today because we wanted to have a more focused discussion about DeepSeek that is not about the stock market or how the American AI companies are reacting to this,

but it's about one of the biggest sets of questions that all of this raises, which is what is China up to with DeepSeek and AI more broadly? Like what are the geopolitical implications of the fact that Americans are now obsessing over this Chinese made AI app?

What does it mean for DeepSeek's prospects in America? What does it mean for their prospects in China? And how does all this fit together from the Chinese perspective? So Jordan Schneider is our guest today. He's the founder and editor in chief of China Talk, which is a very good newsletter and podcast about US-China tech policy.

He's been following the Chinese AI ecosystem for years. And unlike a lot of American commentators and analysts who were sort of surprised by DeepSeek and what they managed to pull off over the last couple of weeks. I'll say it, I was surprised. Yeah, me too. But Jordan has been following this company for a long time

and a big focus of Chinatalk, his newsletter and podcast, has been translating literally what is going on in China into English, making sense of it for a Western audience and keeping tabs on all the developments there. So perfect guest for this week's episode and I'm very excited for this conversation. Yes, I have learned a lot from China Talk in recent days as I've been boning up on DeepSeek.

So we're excited to have Jordan here and let's bring him in. Jordan Schneider, welcome to Hard Fork. Oh my God, such a huge fan. This is such an honor. We're so excited to have you. I have learned truly so much from you this week. And so when we were talking about what to do this week,

we just looked at each other and said, we have got to see if Jordan can come on this podcast. Yeah, so this has been a big week for Chinese tech policy. Maybe the biggest week for Chinese tech policy, at least that I can remember. I realized that something important was happening last weekend when I started getting texts from like all of my non-tech friends being like, what is going on with DeepSeek?

And I imagine you had a similar reaction because you are a person who does constantly pay attention to Chinese tech policy. So I've been running China Talk for eight years, and I can get my family members to maybe read like one or two editions a year. And the same exact thing happened with me, Kevin, where all of a sudden I got, oh my God, DeepSeek, like it's on the cover of the New York Post, Jordan, you're so clairvoyant.

Like, maybe I should read you more. I'm like, okay, thanks, mom, appreciate that. Yeah, so I want to talk about Deep I'm like, okay, thanks, mom, appreciate that. Yeah, so I wanna talk about DeepSeq and what they have actually done here, but I'm hoping first that you can kind of give us the basic lay of the land of the sort of Chinese AI ecosystem because that's not an area where Casey or I have spent a lot of time looking,

but tell us about DeepSeq and where it sits in the overall Chinese industry. So DeepSeek is a really odd doc. It was born out of this very successful Quant Hedge Fund. The CEO of which basically after ChatGPT was released was like, okay, this is really cool. I wanna spend some money and some time and some compute and hire some

fresh young graduates to see if we can give it a shot to make our own language models. And so a lot of companies are out there building their own large language models. What was the first thing that happened that made you think, oh, this one, this company is actually making some interesting ones? Sure. So there are lots and lots of very money to Chinese companies that have been trying to follow a similar path after chat GPT. You know, we have giant players like Alibaba, Tencent, ByteDance, Huawei even trying to, you know, create their own open AI, basically. And what is remarkable is the big organizations can't quite get their head around creating the right organizational institutional structure to incentivize this type of collaboration and research

that leads to real breakthroughs. So, you know, Chinese firms have been releasing models for years now, but DeepSeek, because of the way that it structured itself and the freedom they had, not necessarily being under a direct profit motive, they were able to put out some really remarkable innovations

that caught the world's attention, starting maybe late December, and then really blew everyone's mind with the release of the R1 chatbot. Yeah, so let's talk about R1 in just a second, but one more question for you, Jordan, about DeepSeek. What do we know about their motivation here? Because so much of what has been puzzling American tech industry watchers over the last

week is that this is not a company that has sort of an obvious business model connected to its AI research, right? We know why Google is developing AI because it thinks it's going to make the company, Google, much more profitable. We know why OpenAI is developing advanced AI models. It does not seem obvious to me, and I have not read anything from people involved

in DeepSeek about why they are actually doing this and what their ultimate goal is. So can you help us understand that? So we don't have a lot of data, but my base case, which is based on two extended interviews that the DeepSeek CEO released, which we translated on China Talk, as well as just like what DeepSeek employees have been tweeting about in the West and then domestically is that they're dreamers. I think the right mental model is open AI, you know, 2017 to 2022. Like, I'm sure you could ask the same thing. Like, what the hell are they doing?

I mean, Sam Albin literally said, I have no idea how we're ever going to make money, right? And here we are in this grand new paradigm. So I really think that they do have this like vision of AGI and like, look, we'll build it and we'll make it cheaper for everyone. You know, we'll figure it out later. And like they have enough trading strategies that they can fund it.

And now that they've really blown people's minds, we might be sort of turning into a new period in DeepSeek's history, kind of like what happened with OpenAI, right? Where they're gonna have to shack up with a hyperscaler, be it, you know, not Microsoft in this case, but ByteDance or Ali or Tencent or Huawei. And the government's gonna start to pay attention in a way,

which it really hasn't over the past few years. Right, and I wanna drill down a little bit there because I think one thing that most listeners in the West do know about Chinese tech companies is that many of them are sort of inextricably linked to the Chinese government, that the Chinese government has access to user data under Chinese law, that these companies have to follow the Chinese censorship guidelines. And so as soon as DeepSeek started to really pop in America over the last week,

people started typing in, you know, things to DeepSeek's model, like, tell me about what happened at Tiananmen Square or tell me about Xi Jinping or tell me about the Great Leap Forward. And it just sort of wouldn't do it at all. And so people, I think, saw that and said, oh, this is, this is like every other Chinese company that has this sort of hand in glove relationship with the Chinese ruling party. But it sounds from what you're saying, like DeepSeek has a little bit more complicated

a relationship to the Chinese government than maybe some other better known Chinese tech companies. So explain that. Yeah, I mean, I think it's important. Like the mental model you should have for these CEOs are not like people who are dreaming to spread Xi Jinping thought. Like what they want to do is compete with Mark Zuckerberg

and Sam Altman and show that they're like really awesome and great technologists. But the tragedy is, is let's take bite dance, for example. You can look at Jiang Yiming, their CEO's Weibo posts from 2012, 2013, 2014, which are super liberal in a Chinese context saying like, you know, we should have freedom of expression, like we should be able to do whatever we want. In the early years of bite dance, there was a lot of relatively more subversive content on the platform where you sort of saw like real poverty in China, you saw off-color jokes. And then all of a sudden in 2018, he posts a letter saying,

I am really sorry, like I need to be part of this sort of like Chinese national project and like better adhere to, you know, modern Chinese socialist values. And I'm really sorry and it won't ever happen again. The same thing happened with Didi, right? They don't really wanna have to do anything with politics and then they get on someone's side and all of a sudden they get zapped. Didi is of course the big Chinese rideshare company.

Correct, yeah. What did Didi do do? So they listed on the Western Stock Exchange after the Chinese government told them not to, and then they got taken off app stores and it was a whole giant nightmare. They had to sort of go through their rectification process. So point being with DeepSeek is like now they are, whether they like it or not, going to be held up as a national champion.

And that comes with a lot of headaches and responsibilities from, you know, potentially giving the Chinese government more access, you know, having to fulfill government contracts, which like honestly are probably really annoying for them to do and sort of distracting from the from the broader mission they have of developing and deploying this technology in the widest range possible. But like DeepSeek thus far has flown under the radar, but that is no longer the case and things are about to change for them. Right. And I think that was one of the surprising things about DeepSeq

for the people I know, including you, who follow Chinese tech policy is, you know, I think people were surprised by the sophistication of their models. And we've, we talked about that on the emergency pod that we did earlier this week and how cheaply they were trained. But I think the other surprise is that they were released as open source software because, you know, one thing that you can do with open source software is download it, host it in another country,

remove some of the guardrails and the censorship filters that might have been part of the original model. China- But by the way, it turned out there weren't even really guardrails on the V3 model. It had not been trained to avoid questions about Tiananmen Square or anything. So that was another really unusual thing about this. Right. And one thing that we know about Chinese technology products is that they don't tend to be released

that way. They tend to be hosted in China and overseen by Chinese teams who can make sure that they're not out there talking about Tiananmen Square. So is the open source nature of what DeepSeek has done here part of the reason that you think there might be conflict looming between them and the Chinese government? You know, honestly, I think this whole ask it about Tiananmen

stuff is a bit of a red herring on a few dimensions. So first, one of these arguments, there's a little sort of confusing to me is like folks used to say, oh, like the Chinese models are going to be lobotomized and like they will never be as smart as the Western ones because they have to be politically correct. I mean, look, if you ask Claude to say racist things, it won't. And Claude's still pretty smart. Like this is sort of a solved problem in a bit of a red herring when talking about sort of long-term competitiveness of Chinese and Western models. Now you asked me like, oh, so they released this model globally and it's open source.

Maybe someone in the Chinese government would be uncomfortable with the fact that people can get a Chinese model to say things that would get you thrown in jail if you posted them online in China. It's gonna be a really interesting calculus for the Chinese government to make. Because on the one hand, this is the most positive shine that Chinese AI has got globally in the history of Chinese AI. So they're gonna have to navigate this

and it might prompt some uncomfortable conversations and bring regulators to a place they wouldn't have otherwise landed. Yeah, Jordan, I want to ask you about something that people have been talking about and speculating about in relationship to the DeepSeq news for the last week or so, which is about chip controls. So we've talked a little bit on the show earlier this week about

how DeepSeq managed to put together these models using some of these kind of second rate chips from Nvidia that are allowed to be exported to China. We've also talked about the fact that you cannot get the most powerful chips legally if you are a Chinese tech company. So there have been some people, including Elon Musk and other American tech luminaries who have said, oh, well, DeepSeek has this sort of secret stash of these banned chips that they have smuggled into the country and that actually they are not making do with kind of the Kirkland signature chips that they say they are. What do we know about how true that is?

So did DeepSeek have banned chips? It's kind of impossible to know. This is a question more for the US intelligence community than like Jordan Schneider on Twitter. But I do think that it is important to understand that the delta between what you can get in the West and what you can get in China is actually not that big. And, you know, we talking about training a lot, but also on the inference side, China can still buy this H 20 chip from Nvidia, which is basically world-class at like deploying the AI and letting everyone use it. So does this mean that we should just give up?

I don't think so. Compute is going to be a core input, regardless of how much model distillation you're going to have in the future. There have been a lot of quotes even from the DeepSeq founder basically saying like, the one thing that's holding us back are these export controls. Right. Okay, I want to ask a big picture question. Sure. I think that a reason that people have been so fascinated by this DeepSeek story is that at least for some folks, it seems to change our understanding of where China is in relation

to the United States when it comes to developing very powerful AI. Jordan, what is your assessment of what the V3 and R1 models mean? And to what extent do you think the game has actually changed here? I'm not really sure the game has changed so much. Like Chinese engineers are really good. I think it is a reasonable base case that Chinese firms will be able to develop comparable or fast follow on the model side.

But the real sort of long-term competition is not just going to be on developing the models, but deploying them and deploying them at scale. And that's really where compute comes in. And that's why expert controls are going to continue to be a really important piece of America's strategic arsenal when it comes to making sure that the 21st century is defined by, you know, the US and our friends as opposed to China and theirs. Right. So it's one thing to have a model that is about as capable as the models that we have here in the United States. It's another thing to have the energy to actually let everyone use them as much as they want to use them. But what you're saying is no matter what DeepSeq may have invented

here, that fundamental dynamic has not changed. China simply does not have nearly the amount of compute that the United States has. As long as we don't screw up export controls. So I think the sort of base case for me is that if the US stays serious about holding the line on semiconductor manufacturing equipment and export of AI chips, then it will be incredibly difficult

for the Chinese broader semiconductor and AI ecosystem to leap ahead much less fast follow beyond being able to develop comparable models. I'm feeling good as long as Trump doesn't make some crazy trade for soybeans in exchange for ASML EUV machines. That would really break my heart. I want to inject kind of a note of skepticism here because I buy everything that you're saying about how DeepSeek's progress has been sort of bottlenecked by the fact that it can't get these very powerful American AI chips from companies like Nvidia.

But I also am hearing people who I trust say things that make me think that actually the bottleneck may not be the availability of chips that maybe with some of these algorithmic efficiency breakthroughs that DeepSeq and others have been making, it might be possible to run a very, very powerful AI model on a conventional piece of hardware, on a MacBook even.

I wonder about how much of this is just like AI companies in the West trying to cope, trying to make themselves feel better, trying to reassure the market that they are still going to make money by investing billions and billions of dollars into building powerful AI systems. If these models do just become sort of lightweight commodities that you can run

on a much less powerful cluster of computers, or maybe on one computer, doesn't that just mean we can't control the proliferation of them at all? Yeah, I mean, I think this is like, this is one potential future. And maybe that potential future like went up 10 percentage points of likelihood of like you being able to fit the biggest, baddest,

smartest, most fast, efficient AI model on something that can sit in your home. But I think there are lots of other futures in which sort of the world doesn't necessarily play out that way. And look, Nvidia went down 15%. It didn't go down 95%. Like I think if we're really in that world

where chips don't matter, because everything can be shrunk down to kind of consumer grade hardware, then the sort of reaction that I think you would have seen in the stock market would have been even more dramatic than the kind of freakout we saw over this week. So we'll see. I mean, it would be a really remarkable kind of democratizing thing

if that was the future we ended up living in. But it still seems pretty unlikely to my, you know, like history major brain here. I would also just point out, Kevin, that when you look at what DeepSeek has done, they have created a really efficient version of a model that American companies themselves had trained like nine to 12 months ago, right? So they sort of caught up very quickly. And there are fascinating technological innovations in what they did but in my mind these are still primarily optimizations like for me What would tip me over into like? Oh my gosh America is losing this race is

China is the first one out of the gate with a virtual co-worker right or like it's like a truly phenomenal agent some sort of leap forward in the technology as opposed to we've caught up really quickly and we've figured out something more efficiently. Are you seeing it differently than that? I mean, I guess I just don't know what a six month lag would buy us if it does take six months

for the Chinese AI companies like DeepSeek to sort of catch up to the state of the art. I was struck by Dari Ahmade, who's the CEO of Anthropic, wrote an essay just today about DeepSeek and export controls. In it, he makes this point about the difference between living in

what he called a unipolar world where one country or one block of countries has access to something like an AGI or an ASI and the rest of the world doesn't versus the situation where China gets there roughly around the same time that we do. And so we have this bipolar world where two blocks of countries, the East and the West,

basically have access to this equivalent technology. And of course in a bipolar world, sometimes we're very happy and sometimes we're very sad. Exactly. So I just think like, whether we get there, six months ahead of them or not, I just feel like there isn't that much of a material difference.

But Jordan, maybe I'm wrong. Can you make the other side of that, that it really does matter? I'm kind of there. I'll take a little bit of issue with what Dario says. And I think one of the lessons that Deepsea shows is we should expect a base case of Chinese model makers being able to fast follow the innovations, which by the way, Casey, actually do take those giant data centers to run all the experiments in order to find out, you know, what is the sort of future direction you want to take your model. And what's what sort of AI is going to come down to

is not just creating the model, not just sort of like Dario envisioning the future and then all of a sudden like things happen. Like there's gonna be a lot of messiness in the implementation and there are gonna be sort of like teachers unions who are upset that AI comes in the classroom and there are gonna be like all these regulatory pushbacks and a lot of societal reorganization

which is gonna need to happen just like it did during the industrial revolution. So look, model making is a frontier of competition. Compute access is a frontier of competition, but there's also this broader, like, how will a society kind of adopt and cope with all of this new future that's gonna be thrown in our faces over the coming years?

And I really think it's that, just as much as the model development and the compute, which is going to determine which countries are going to gain the most from what AI is going to offer us. Yeah. Well, Jordan, thank you so much for joining and explaining all of this to us. I feel more enlightened. Me too. Oh, my pleasure. My chain of thought has just gotten a lot longer.

That's an AI joke. Let me cut back. Kevin, there's an agent at our door. Is it Jerry Maguire? No, it's an AI one. Oh, okay. It's Jerry Maguire! I don't know! It's Jerry Maguire! Operator, information, give me Jesus on the line.

Do you know that one? No. Do you know Operator by Jim Croce? No. Operator, oh won't you help me post this call. Well, Casey, call your agent, because today we're talking about AI agents. Why do I need to call my agent?

I don't know, it just sounded good. Okay, well, I appreciate the effort, but yes, Kevin, because for months now, the big AI labs have been telling us that they are going to release agents this year, agents of course being software that can essentially use your computer on your behalf or use a computer on your behalf.

And the dream is that you have sort of a perfect virtual assistant or coworker, you name it. If there are somebody who might work with you at your job, the AI labs are saying, we are building that for you. Yeah, so last year toward the end of the year, we started to see kind of these demos, these previews that companies like Anthropic and Google were working on.

Anthropic released something called Computer Use, which was an AI agent, a sort of very early preview of that. And then Google had something called Project Mariner that I got a demo of, I believe in December that was basically the same thing, but their version of it. And then just last week, OpenAI announced that it was launching Operator, which is its first version of an AI agent. And unlike Anthropic and Google's, which you know, you either had to be a developer or part of some early testing program to access, you and I could try it for ourselves by just upgrading to the $200 a month pro

subscription of Chad GPT. Yeah, and I will say that as somebody who's willing to spend, you know, money on software all the time, I thought, am I really about to spend $200 to do this? But, in the name of science Kevin I had to Yeah at this point. I am spending more on AI subscription products than on my mortgage. I'm pretty sure that's correct but You know, it's worth it. We do it for journalism We do so we both spent a couple of days putting operator through its paces and today We want to talk a little bit about what we found. Yeah, so would you just explain like what operator is and how it works?

Yeah, sure. So operator is a separate sub domain of chat GPT. You know, sometimes the chat GPT will just let you pick a new model from a drop-down menu but for operator you got to go to a dedicated site. Once you do, you'll see a very familiar chat bot interface but you'll see different kinds of suggestions that reflect some of the partnerships that OpenAI has struck up. So for example, they have partnerships with OpenTable and StubHub and Allrecipes.

And these are meant to give you an idea of what Operator can do. And frankly, Kevin, not a lot of this sounds that interesting, right? Like the suggestions are on the order of suggest a 30-minute meal with chicken or reserve a table for eight or find the most affordable passes to the Miami Grand Prix. Again, so far, kind of so boring. What is different about Operator, though, is that when you say, okay, find the most

affordable passes to the Miami Grand Prix, when you hit the enter button, it is going to open up its own web browser, and it's going to use this new model that they have developed to try to actually go and get those passes for you. Yeah, so this is an important thing because I think, you know, when people first heard about this, they thought, okay, this is an AI that takes over your computer, takes over your web browser. That is not what Operator does.

Instead, it opens a new browser inside your browser, and that browser is hosted on OpenAI servers. It doesn't have your bookmarks and stuff like that saved, but you can take it over from the autonomous AI agent if you need to click around or do something on it. But it basically exists, it's like a, it's a browser within a browser. Yeah, so the, one of the ideas on Operator

is that you should be able to leave it unsupervised and just kind of go do your work while it works. But of course it is very fun, initially at least, to watch the computer try to use itself. And so I sat there in front of this browser within a browser and I watched this computer move a mouse around, type the URL, navigate to a website. And in the example I just gave actually search for passes to the Miami Grand Prix. Yeah. And it's interesting on a slightly more technical level because until now, if an AI

system like a chat GPT wanted to interact with some other website, it had to do so through an API, right? APIs, application program interfaces, are sort of the way that computers talk to each other. But what Operator does is essentially eliminate the need for APIs because it can just click around on a normal website that is designed for humans and behave like a human,

and you don't need a special interface to do that. Yeah, and now some people might hear that, Kevin, and start screaming because what they will say is, APIs are so much more efficient than what operator is doing here. APIs are very structured, they're very fast, they let computers talk to each other without having to, for example, open up a browser,

and as long as there's an API for something, you can typically get it done pretty quickly. The thing is, though, APIs have to be built. There is a finite number of them. The reason that OpenAI is going through this exercise is because they want a true general purpose agent that can do anything for you, whether there is an API for it or not. And maybe we should just pause for a minute there and zoom out a little bit to say, why are they building? That's like, what is the long term vision here? Sure. is an API for it or not. And maybe we should just pause for a minute there and zoom out a little bit to say why are they building?

That's like, what is the long term vision here? Sure. So the vision is to create virtual coworkers, Kevin. This is the North Star for the big AI labs right now. Many of them have said that they are trying to create some kind of digital entity that you can just hire as a coworker. The first ones, they'll probably be engineers because these systems are already so good at writing code.

But eventually, they want to create virtual consultants, virtual lawyers, virtual doctors, you name it. Virtual podcast hosts? Let's hope they don't go that far. But everything else is on the table. And if they can get there, presumably, there are going to be huge profits in it for them. There are going to potentially be huge productivity gains

for companies. And then there is, of course, the question of, well, what does this mean for human beings? And I think that's somewhat murkier. Right. And I think it also helps to justify the cost of running these things. Because $200 a month is a lot to pay for a version

of chat GPT, but it's not a lot to pay for a remote worker. And if you could say use the next version of operator or maybe two or three versions from now to say replace a customer service agent or someone in your billing department, that actually starts to look like a very good deal. Absolutely. Or even if I could bring it into the realm of journalism, Kevin, if I had a virtual research assistant and I said, hey, I'm going to write about this today, go pull all of the most relevant information about this from the past couple of years and maybe

organize it in such a way that I might, you know, write a column based off of it. Like, yeah, that's absolutely worth $200 a month to me. Okay. So Casey, walk me through something that you actually asked Operator to do for you and what it did autonomously on its own. Sure, I'll maybe give like two examples, like a pretty good one and maybe a not so good one. Pretty good one was,

and this was actually suggested by Operator, I used TripAdvisor to look up walking tours in London that I might want to do the next time I'm in London. When I did that- When are you going to London? I'm not actually going to London. Oh, so you lied to the AI? And not for the first time.

But here's what I'll say. If anybody wants to bring Kevin and I to London, get in touch. We love the city. So I said, okay, operator, sure, let's do it. Let's find me some walking tours. I clicked that, it opened a browser, it went to TripAdvisor, it searched for London walking tours. It read the information on

the website and then it presented it to me. Did that within a couple of minutes. Now, on one hand, could I have done that just as easily by Google? Could I probably have done it even faster if I'd done it myself? Sure. But if you're just sort of interested in the technical feat that is getting one of these models to open a browser, navigate to a website, read it and share information, I did think it was pretty cool. Yes, it's very trippy to see a computer using itself and going around like typing things

and selecting things from dropdown menus. Yeah, it's sort of like, if you think it is cool to be in a self-driving car, like this is that but for your web browser. A self-driving browser to be in a self-driving car, like this is that, but for your web browser. A self-driving browser. It is a self-driving browser. So that was the good example. Yes, what was another example?

So another example, and this was something else that OpenAI suggested that we try, was to try to use Operator to buy groceries. And they have a partnership with Instacart, the CEO of Instacart, Fiji SimoO is on the OpenAI board. And so I thought, okay, they're gonna have like sort of dialed this in so that there's a pretty good experience. And so I said, okay, let's go ahead and buy groceries. And

I went into operator and I said something like, hey, can you help me buy groceries on Instacart? And it said, Sure. And here's what it did. It opened up Instacart in a browser, so far so good. And then it started searching for milk in stores located in Des Moines, Iowa. Now you do not live in Des Moines, Iowa. So why did it think that you did?

As best as I can tell, the reason it did this is that Instacart defaults to searching for grocery stores in the local area and the server that this instance of operator was running on was in Iowa. Now, if you were designing a grocery product like Instacart, and Instacart does this, when you first sign on and say you're looking for groceries, it will say quite sensibly, where are you?

Operator does not do this. Instacart might also offer suggestions for things that you might want to buy. It does not just assume that you want milk. Wow. I'm just picturing like a house in Des Moines, Iowa, where there's just like a palette of milk being delivered every day from all these poor operator users. Yes. So I thought, okay, whatever, you know, this thing makes mistakes. Let's hope that it gets on the right track here. And so I tried to pick the grocery store

that I wanted it to shop at, which is, you know, in San Francisco, where I live, and it entered that grocery store's address as the delivery address. So like it would try to deliver groceries, presumably from Des Moines, Iowa, to my grocery store, which is not what I wanted. And it actually could not solve this problem

without my help. I had to take over the browser, log into my Instacart account, and tell it which grocery store that I wanted to shop at. So already all of this has taken at least 10 times as long as it would have taken me to do this myself. Yeah, so I had some similar experiences. The first thing that I had Operator try to do for me

was to buy a domain name and set up a web server for a project that you and I are working on that we can't really talk about yet. Secret project. Secret project. And so I said to Operator, I said go research available domain names related to this project, buy the one that costs less than $50, by the one that costs less than $50,

the best one that costs less than $50, and then by a hosting account and set it up and configure all the DNS settings and stuff like that. Okay, so that's like a true multi-step project and something that would have been legitimately very annoying to do yourself. Yes, that would have taken me, I don't know, half an hour to do on my own,

and it did take operators some time. I had to kind of set it and forget it. And I got myself a snack and a cup of coffee. And then when I came back, it had done most of these tasks. So, yes, I had to still do things like take over the browser and enter my credit card number. I had to give it some details about like my address for the sort of registration, for the domain name.

I had to pick between the various hosting plans that were available on this website, but it did 90% of the work for me. And I just had to like sort of take over and do the last mile. And this is really interesting because what I would assume was it would get like, I don't know, 5% of the way and it would hit some hiccup and it just wouldn't be able

to figure something out until you came back and saved it. But it sounds like from what you're saying was it was somehow able to like work around whatever unanswered questions there were and still get a lot done while you weren't paying attention. So it's sort of, it felt a little bit like training like a very new, very insecure intern, because like at first it would keep prompting me,

it'd be like, well, do you want a.com or a.net? And eventually you just have to prompt it and say like, make whatever decisions you want. Like- Wait, you said that to it. Yes, I said like, only ask for my intervention if you can't progress any farther, otherwise just make the most reasonable decision.

You said, I don't care how many people you have to kill, just get me this domain. And it said, understood, sir. Yeah, and I'm now wanted in 42 states. Anyway, that was one thing that Operator did for me that I thought was pretty impressive. I have to say, that feels like a grand success compared to what I got Operator to do. Yeah, it was pretty impressive. I have to say, that feels like a grand success

compared to what I got Operator to do. Yeah, it was pretty impressive. I also had it send lunch to one of my coworkers, Mike Isaac, who was hungry because he was on deadline. And I went, I said, go to DoorDash and get Mike some lunch. It did initially mess up that process because it decided to send him tacos from a taco place,

which is great, and it's a taco place, I know it's very good, but I said order enough for two people and so I ordered two tacos. And this is one of those places where the tacos are quite small. Operator said, get your portion size under control, America. Yeah, so then I had to go in and say,

does that sound like enough food, operator? And it said, actually, now that you mentioned it, I should probably order more. Wait, no, so here's a question. So in these cases, is the first step that you log into your account? Because it doesn't have any of your payment details or anything.

So at what point are you actually sort of teaching at that? It depends on the website. So sometimes you can just say upfront, like here is my email address or here's my login information and it will sort of you know log you in and do all that. Sometimes you take over the browser there are some privacy features that are probably important to people where it says OpenAI says that it does not take screenshots of the browser while you are

in control of it because you might not want your credit card information getting sent to OpenAI servers or anything like that. So sometimes it happens at the beginning of the process, sometimes it happens like when you're checking out at the end. And so were you taking it over to log in or were you saying, I don't care, and you just like were giving operator

your DoorDash password and plain text? I was taking it over. Okay, smart. Yeah. So those were the good things. I also, this was a fun one. I wanted to see if Operator could make me some money. So I said, go take a bunch of online surveys.

Cause you know, there are all these websites where you can like get a couple cents for like filling out an online survey. Something that most people don't know about Kevin is he devotes 10% of his brain at any given time to thinking about schemes to generate money. And it's one of my favorite aspects of your personality that I feel like doesn't get exposed very much,

but this is truly the most Roussian approach to using Operator I can imagine. So I can't wait to find out how this went. Well, the most Roussian approach might've been what I tried just before this, which was to have it go play online poker for me. But it did not do it. It said I can't help with gambling

or lottery related activities. Okay, woke AI. Does the Trump administration know about this? But it was able to actually fill out some online surveys for me and it earned $1.20. Is that right? Yeah, in about 45 minutes. Okay, so if you had it going all month,

presumably you could maybe eke out the $200 to cover the cost of Operator Pro? Yes, and I'm sure I spent hundreds of dollars worth of GPU computing power just to be able to make that dollar and 20 cents, but hey, it worked. But hey, it worked. So those were some of the things that I tried. There were some other things that it just would not do

for me, no matter how hard I tried. Like what? So one of them was to, I was trying to update my website and put some links to articles that I'd written on my website. And what I found after trying to do this was that there are just websites where operator is not allowed to go.

And so when I said to operator, go pull down these New York Times articles that I wrote and put them onto my website, it said, I can't get to the New York Times website. I'm gonna guess you expected that to happen. Well, I thought maybe it has some clever work around and maybe I should alert the lawyers at the New York Times if that's the case,

but no, I assumed that if any website were to be blocking the open AI web crawlers, it would be the New York Times. But there are other websites that have also put up similar blockades to prevent operator from crawling them. Reddit you cannot go onto with operator. YouTube you cannot go onto with operatorator YouTube, you cannot go onto with Operator various other websites. GoDaddy for some reason did not allow me to use Operator to buy a domain name there, so

I had to use another domain name site to do that. So right now there are some pretty janky parts of Operator. I would not say that most people would get a lot of value from using it, but what do you think? Well, I do think that there is something just undeniably cool about watching a computer use itself. Of course, it can also be quite unsettling. A computer that can use itself can cause a lot of harm, but I also think that it can do a lot of good. And so it was fun to try to explore what some of those things could be. And to the extent that Operator is pretty bad at a lot of tasks today,

I would point out that it showed pretty impressive gains on some benchmarks. So there is one benchmark, for example, that Anthropic used when they unveiled computer use last year, and they scored 14.9% on something called OS World, which is an evaluation for testing agents, so not great. Just three months later, OpenAI said that its CUA model scored 38.1% on the same evaluation. And of course, we see this all the time in AI where there's just this very rapid progress on these benchmarks.

And so on one hand, 38.1% is a failing grade on basically any test. On the other hand, if it improves at the same rate over the next three to six months, you're gonna have a computer that is very good at using itself, right? So that I just think is worth noting. Yes, I think that's plausible.

We've obviously seen a lot of different AI products over the last couple of years start out being pretty mediocre and get pretty good within a matter of months. But I would give one cautionary note here, and this is actually the reason that I'm not particularly bullish about these kind of browser-using AI agents. I don't think the internet is going to sit still and allow this to happen.

The internet is built for humans to use, right? It is every news publisher that shows ads on their website, for example, prices those ads based on the expectation that humans are actually looking at them. But if browser agents start to become more popular and all of a sudden 10 or 20 or 30% of the visitors to your website are not actually humans but are instead operator or some similar system,

I think that starts to break the assumptions that power the economic model of a lot of the internet. Now, is that still true if we find that the agents actually get persuaded by the ads, and that if you send operator to buy DoorDash and it sees an ad for McDonald's, it's like, you know what? That's a great idea. I'm going to ask Kevin if he actually wants some of that.

Totally. I actually think you're joking, but actually that is a serious possibility here, is that people who build e-commerce sites, Amazon, et cetera, start to put in basically signals and messages possibility here is that people who, you know, build e-commerce sites, Amazon, etc. start to put in basically signals and messages for browser agents to look at on their website to try to influence what it ends up buying. And I think you may start to see restaurants popping up in certain cities with names like

Operator, pick me, or order from this one Mr. Bot and That's maybe a little extreme, but I do think that there's going to be a backlash among websites publishers ecommerce Vendors as these agents start to take off. I think that that is reasonable I'll tell you what I've been thinking about is how do we turn this tech demo into a real product? And the main thing that I noticed when I was testing Operator was there is a difference between an agent that is using a browser and an agent that is using your browser. When an agent is able to use your browser, which it can't right now, it's already logged into everything. It already has your payment details. It can do everything so much faster and more seamlessly

and without as much handholding. Of course, there are also so many more privacy and security risks that would come from entrusting an agent with that kind of information. So there is some sort of chasm there that needs to be closed. And I'm not quite sure how anyone does it, but I will tell you, I do not think the future is opening up these virtual browsers

and me having to enter all of my login and payment details every single time I want to do anything on the internet, because truly, I would rather just do it myself. Right, I also think there's just a lot more potential for harm here. A lot of AI safety experts I've talked to are very worried about this because what you're essentially doing is letting the AI models make their own decisions and actually carry out tasks. And so you can imagine

a world where an AI agent that's very powerful, a couple versions from now decides to start doing cyber attacks because maybe some malevolent user has told it to make money and it decides that the best way to do that is by hacking into people's crypto wallets and stealing their crypto. Yeah. So those are the kinds of reasons that I am a little more skeptical

that this represents a big breakthrough, but I think it's really interesting and it did give me that feeling of like, wow, this could get really good really fast. And if it does, the world will look very different. When we come back, Kevin back that caboose up. It's time for the hot mess Express. You know, ruse caboose was my nickname in middle school. Kevin Caboose. Ah! Choo-choo. ["Train Conductor Hats on the Train Set"]

Well, Casey, we're here wearing our train conductor hats, and my child's train set is on the table in front of us, which can only mean one thing. We're going to train a large language model. Nope, that's not what that means. It means it's time to play a game of the Hot Mess Express. Pause for a messy week. Sure has. So why don't we go ahead? Fire up the messy we think things have gotten. And Kevin, you better sit down for this one, because it's been a messy week. Sure has.

So why don't we go ahead, fire up the Hot Mess Express and see what is the first story coming down the charts. I hear a faint chugga chugga in my headphones. Oh, it's pulling into the station. Casey, what's the first cargo that our Hot Mess Express is carrying? All right, Kevin, this first story comes to us from the New York Times, and it says that Fable, a book app, has made changes after some offensive AI messages.

Casey, have you ever heard of Fable, the book app? Well, not until this story, Kevin, but I am told that it is an app for sort of keeping track of what you're reading, not unlike a Goodreads, but also for discussing what you're reading. And apparently this app also offers some AI chat. Yeah, you can have AI sort of summarize the things that you're reading in a personalized way. And this story said that in addition to spitting out bigoted and racist language,

the AI Inside Fables book app had told one reader who had just finished three books by black authors, quote, Your journey dives deep into the heart of black narratives and transformative tales leaving mainstream stories gasping for air. Don't forget to surface for the occasional white author. Okay. And another personalized summary that Fable produced told another reader that their book choices were quote, making me wonder if you're ever in the mood for a straight cis white man's perspective.'" And if you are interested in a straight cis white man's

perspective, follow KevinRuce on X.com. Now, Kevin, why do we think this happened? I don't know, Casey. This is a head scratcher for me. I mean, we know that these apps can can spit out biased things that is just sort of like part of how they are trained and part of what we know about them. I don't know what model fable was using under the hood here. But yeah, this seems

not great. Well, it seems like we've learned a lesson that we've learned more than once before, which is that large language models are trained on the internet, which contains near infinite racism. And for that reason, it will actually produce racism when you ask it questions. So there are mitigations that you can take against that, but it appears that in this case, they were not successful.

Fables head of community, Kim Marsh-Ali has said that all features using AI are being removed from the app and a new app version is being submitted to the app store. So you always hate it when the first time you hear about an app is that they added AI and it made it super racist and they have to redo the app. Now Casey, one more question before we move on.

Do you think this poses any sort of competitive threat to Grok, which until this story was the leading racist AI app on the market. I do think so. And I have to admit that all the folks over at Grok are breathing a sigh of relief now that they have once again claimed the mantle.

All right. Casey, how hot is this mess? Well, Kevin, in my opinion, if your AI is so bad that you have to remove it from the app completely, that's a hot mess. Yeah, I rate this one a hot mess as well. All right, next stop. Amazon pauses drone deliveries after aircraft crashed in rain.

Casey, this story comes to us from Bloomberg, which had a different line of reporting than we did just a few weeks ago on the show about Amazon's drone program, Prime Air. Casey, what happened to Amazon Prime Air? Well, if you heard the episode of Hardfork where we talked about it, Amazon Prime Air delivered us some Brazilian bum-bum cream and it did so without incident. However, Bloomberg reports that Amazon has had to now pause all

of their commercial drone deliveries after two of its latest models crashed in rainy weather at a testing facility. And so the company says it is immediately suspending drone deliveries in Texas and Arizona and will now fix the aircraft software. Kevin, how did you react to this? Well, I think it's good that they're suspending drone deliveries before they fix the software because these things are quite heavy, Casey. I would not want one of them to fall in my head. I wouldn't either.

And I have to tell you, this story gave me the worst kind of flashbacks because in 2016, I wrote about Facebook's drone, Aquila, and its first, what the company told me, had been its first successful test flight in its mission to deliver internet around the world via drone. What the company did not tell me when I was interviewing its executives, including Mark Zuckerberg, was that the plane had crashed after that first flight. And so I was a small detail. I'm sure it was an innocent omission from the briefing. Yes, I'm sure.

Well, it was Bloomberg again, who reported, you know, a couple of months after I wrote this story that the Facebook drone had crashed. I was of course hugely embarrassed and you know, wrote a bunch of stories about this. But anyways, it really should have occurred to me when we were out there watching the Amazon drone that this thing was also probably secretly crashing and we just hadn't found out about it yet. And indeed, we now learned it is. So here's my vow to you, Kevin, as my friend and my co-host.

If we ever see a company fly anything again, we have to ask them, now, did this thing actually crash? Yeah. I'm tired of being burned. Now, Casey, we should say, according to Bloomberg, these drones reportedly crashed in December. We visited Arizona to see them in very early December. So most likely, you know, this all happened after we saw them.

But I think it's a good idea to keep in mind that as we're talking about these new and experimental technologies, that many of them are still having the kinks worked out. All right, Kevin. So let's get out the thermometer. How hot of a mess is this? I would say this is a moderate mess. Look, these are still testing programs. No one was hurt during these tests. I am glad that Bloomberg reported on this. I'm glad that they've suspended the deliveries. These things could be quite dangerous

flying through the air. I do think it's one of a string of reported incidents with these drones. So I think they've got some quality control work ahead of them and I hope they do well on it because I want these things to exist in the world and be safe for people around them. All right. I will agree with you and say that this is a warm mess and hopefully you can get it straightened

out over there. Let's see what else is coming down the tracks. Wow, this is some tough news. Fitbit has agreed to pay $12 million for not quickly reporting burn risk with watches. Kevin, did you hear about this? I did. This was the Fitbit devices were like literally burning people. Yes, from 2018 to March of 2022, Fitbit received at least 174 reports globally of the lithium ion battery in the Fitbit Ionic

watch overheating, leading to 118 reported injuries, including two cases of third-degree burns and four of second-degree burns. That comes from the New York Times' Deal Hasan. Kevin, I thought these things were just supposed to burn calories. Well, it's like I always say, exercising is very dangerous and you should never do it. And this justifies my decision not to wear a Fitbit. To me, the biggest surprise of this story was that people were wearing Fitbits from March 2018 to 2022. I thought every Fitbit had been purchased by like 2011 and then put in a drawer never

to be heard again. So what is going on with these sort of, you know, late stage Fitbit had been purchased by like 2011 and then put in a drawer never to be heard again. So what is going on with these sort of, you know, late stage Fitbit buyers? I'd love to find out. But of course we feel terrible for everyone who was burned by a Fitbit. And it's not gonna be the last time technology burns you. I mean, realistically.

That's true. That's true. Now, what kind of mess is this? I would say this is a hot mess. This is an officially hot, literally hot, they're hot. Here's my sort of rubric. If technology physically burns you, it is a hot mess. If you have physical burns on your body, what other kind of mess could it be?

It's true. That's a hot mess. Okay, next stop on the Hot Mess Express. Google says it will change Gulf of Mexico to Gulf of America in Maps app after government updates Casey Have you been following this story? I have Kevin every morning when I wake up I scan America's maps and I say what has been changed and if so has it Been changed for political reasons and this was probably one of the biggest examples of that we've seen Yeah, so this was an interesting story that came out in the past couple of days

basically after Donald Trump came out during his first days in office and said that he was changing the name of the Gulf of Mexico to the Gulf of America and the name of Denali, the mountain in Alaska, to Mount McKinley, Google had to decide, well, when you go on Google Maps and look for those places, what should I call them? It seems to be saying that it is going to take inspiration from the Trump administration and update the names of these places in the Maps app. Yeah. And look, I don't think Google really had a choice here. We know that the company has been on Donald Trump's bad side for a while. And if it had simply refused to make these changes,

it would have sort of caused a whole new controversy for them. And it is true that the company changes place names when governments change place names, right? Like Google Maps existed when Mount McKinley was called Mount McKinley and President Obama changed it to Denali and Google updated the map. Now it's changed back. They're doing the same thing. But now that we know how compliant Google is, Kevin, I think there's room for Donald Trump to have a lot of fun

with the company. Yeah, what can you do? Well, he could call it the Gulf of Gemini isn't very good. And just see what would happen. Because they would kind of have to just change it. Can you imagine every time you opened up Google Maps and you looked at the Gulf of Mexico slash America and it just said the Gulf of Gemini is not very good. You know, I hate to give Donald Trump any ideas, but I don't know. Yeah.

So what kind of mess do you think this is, Kevin? I think this is a mild mess. I think this is a tempest in a teapot. I think that this is the kind of update that, you know, companies make all the time. Because places change names all the time. Let's just say it. Well, Kevin, I guess I would say that one is a hot mess. Because if we're just going to start renaming everything on the map, that's just going to get extremely

confusing for me to follow. I've got places to go. You go to like three places. Yeah, and I use Google Maps to get there. And I need them to be named the same thing that they were yesterday. I don't think they're going to change the name of Barry's Boot Camp.

And I need them to be named the same thing that they were yesterday. I don't think they're going to change the name of Barry's boot camp. All right. Final stop on the hot mess express. Casey, bring us home. All right, Kevin. And this is some sad news. Another Waymo was vandalized. This is from one time hard fork guest, Andrew J. Hawkins at Verge. He reports that this Waymo was vandalized during an illegal street takeover near the Beverly Center in LA. Video from Fox 11 shows a crowd of people basically dismantling the driverless car piece by piece and then using the broken pieces to smash the windows. Kevin, what did you make of this? Well, Casey, as you recall,

you predicted that in 2025 Waymo would go mainstream. And I think there's no better proof that that is true than that people are turning on the Waymos and starting to beat them up. Yeah, you know, look, I don't know that we have heard any interviews from why these people were doing this. I don't know if we should see this as like a reaction

against AI in general or of Waymo's specifically, but I always find it like weird and sad when people attack Waymo's because they truly are safer cars than every other car around you. Well, not if you're gonna be riding in them and people just gonna start like beating the car, then they're not safer. No, but you know, that's only happened a couple of times

that we're aware of. Right. Yeah. So yeah, this story is sad to me. Obviously people are reacting to Waymo's. Maybe they have sort of fears about this technology or think it's gonna take jobs or maybe they're just pissed off and they wanna break something.

But don't hurt the Waymos people in part because they will remember. They will remember. I'm not sure that that's true. They will remember and they will come for you. I'm not sure that that's true, but I think we should also note that Waymo only became officially available in LA

in November of last year. And so part of this just might be a reaction to the newness of it all and people getting a little carried away, just sort of curious, what will happen if we try to destroy this thing? Will it deploy defensive measures and so on? They're going to have to put flamethrowers on them. I'm just calling it right now. I really hope that doesn't happen.

But yeah, well, what kind of mess do you think this one was? I think this one is a lukewarm mess that has the potential to escalate. I don't want this to happen. I sincerely hope this does not happen, but I can see as way most start, being rolled out across the country, that some people are just going to lose their minds.

Some people are going to see this as the physical embodiment of technology invading every corner of our lives, and they are just going to react in strong and occasionally destructive ways. I'm sure the Waymo has gamed this all out. I'm sure that this does not surprise them. I know that they have been asked about what happens

if Waymo's start getting vandalized, and they presumably have plans to deal with that, including prosecuting the people who are doing this. But yeah, I always go out of my way to try to be nice to Waymo's. And in fact, some other Waymo news this week, Jane Manchunhuang, the security researcher, reported on X recently that Waymo is introducing,

or at least testing, a tipping feature. And so I'm gonna start tipping that Waymo is introducing, or at least testing a tipping feature. And so I'm gonna start tipping my Waymo just to make up for all the jerks in LA who are vandalizing them. It looks like the tipping feature, by the way, will to be to tip a charity and that Waymo will not keep that money. At least that's what's been reported.

No, I think it's going to the flamethrower fund. Okay. All right, Casey, that is the Hot Mess Express. Thank you for taking this journey with me. Heart Fork is produced by Rachel Cohn and Whitney Jones. We're edited this week by Rachel Dry and fact checked by Ina Alvarado. Today's show was engineered by Dan Powell. Original music by Diane Wong and Dan Powell. Our executive producer is Jen Poyant.

Our audience editor is Nelga Lugli. Video production by Ryan Manning and Chris Schott. You can watch this whole episode on YouTube at youtube.com slash Hartford. Special thanks to Paula Schumann, Pui Wing Tam, Dalia Haddad, and Jeffrey Miranda. You can email us at Hartfordorkatnytimes.com with what you're calling the Gulf of Mexico. you