Big Tech Earnings Analysis, Media Startup Fact-Checking AI, and Grammarly’s Rebrand | Oct 30, 2025

25:47It was as simple as that. We didn't anticipate AI. We became the first firm a year after we were created to say we wanted to invest in AI. That was over my objection at the time. I was completely wrong. I just thought it was too early. And so we've been doing that since 2014. We made our first generative AI investment in 2016. And the work experience is still terrible for many people. They don't like the experience of being at work. They don't feel as productive as they could be. Many people in the economy can't access work that pays them enough, even if they want it. And so as far as I'm concerned, we have enormous investment opportunities.

26:22And I think the range of companies we've been involved in. I could have never guessed that one of the most successful investments we've ever made would be at a shipping broker. You mentioned FlexCord. We got Masterclass, Weights and Biases, which just sold to CoreWeave, Repli. I guess the point I'm just making is up and down the stack of possibilities, there's still a lot we can make better about work. RAOUL PAL So on that topic of work, then, where do you stand on this issue of, hey, to what extent is AI going to be able to do people's jobs? What's going to happen with all of this labor and talent, is it going to get replaced?

26:54Where do you stand on that? RAOUL PALERMOUSH - Yeah. So first of all, I think it's not - RAOUL PALERMOUSH - And maybe before you get, what I would love to get to, Roy, is I ask this question to people and people say, well, it's going to be somewhere in the middle. And my follow-up question is - ROOUL PALERMOUSH - It's not going to be in the middle. A hundred percent of jobs will be replaced. RAOUL PALERMOUSH - Okay. So what are people going to do then? RAOUL PALERMOUSH - Yeah. So first of all, I don't think it's the right question, just to say. I think the better question is, in a world with AI, just what will work look like in general?

27:20What kinds of things will people do, which is what you were just getting at in your call? The reason I think that's a better question is because AI is a mix of what I call a loom and a crane, two different kinds of technologies. A spinning loom is a technology designed to replace a human weaver. It first augments them by helping them weave faster, but eventually replaces them. A crane, like a construction crane, is a piece of technology that doesn't automate anything. It allows human beings to do something that in the absence of a crane, they could never otherwise do. and AI is both of those at the same time.

27:51And that's why, and we tend to imagine replacement because it's just easier to imagine. Eric Brynjolfsson, the economist, we also backed his startup, calls it the Turing trap. And simple answer, the Dodge answer, but it's simple and we can talk more, is it'll replace 100 % of jobs because technology always has. If our great-grandparents saw what you and I are doing right now, I don't know where your great-grandparents were. My grandfather lost his eye working in a mine. They'd be like, they're not working. This looks like fun. Nobody's going to get physically harmed. But I think in the West, and particularly in the United States, we redefine whatever it is we do with our days as work.

28:28Figure it out. Figure it out. And again, I think the real question, though, the challenge is, I'll give you the precise version of it is, what's a job that millions of people can do that will allow them to feed their families that doesn't yet exist or doesn't yet exist in the scale of millions of people? And that's where I just don't know the answer. Will home healthcare aid feed millions of people? I don't know. Will vibe... Dot, dot, dot. Who knows? Dot, dot, dot. Well, I think that's a good place to wrap it up. Thank you, Roy, for coming on the show. We really appreciate it. Some great insights.

29:00We'll have you on soon. I really appreciate you guys doing this and you having me. Take care. Of course. That's Roy Bott from Bloomberg Beta. Okay. Grammarly is getting a new name. The AI powered writing assistant is rebranding as Superhuman, merging Grammarly with two of its acquired companies, Coda, an all-in-one productivity platform, and Superhuman Mail, an AI-driven email app. Joining me to discuss the company's strategy are Shishir Mehrotra, CEO of Superhuman, and Rahul Vora, head of Mail. Welcome to the both of you. It's great to have you. Thank you. Great to be here. Thanks for having us.

29:34And I should say, Shishir, it's your second time on the show. Welcome back, Rahul. It's your first time. So we're excited to have you here. And I'm excited to talk about the new name. By the way, before we get into the new name, Wasn't this a prediction, Shishir, that people made that when you bought Superhuman, there was some chatter online saying, oh, this is going to, I mean, you know, people were calling this at the outset. Well, I mean, I think there were definitely a lot of suggestions for it. I think it's a great name. So it's not a, it's maybe a surprising pattern because it's not very common to do the elevation of a sub-brand into the corporate brand.

30:06But it's a fantastic name. We also teased it a bit. And we wanted to, we can talk more about that. It actually turns out to be really logistically hard to do a brand swap this way. And so we wanted to we wanted to get the news out there a little bit. Hard, hard. Why? Tell us about that. Oh, boy. This, you know, it's funny. I think I think this analogy I was using with the team naming a company super hard. It's kind of like naming a kid. Renaming a 16 year old company is like 10x is hard. It's like renaming a 16 year old kid and swapping the name of your 16 year old and your 11 year old, like 100x is hard.

30:41And it means things like you're not only changing a corporate name, you know, if you think about other cases where people have done that, you know, Alphabet and Google or Meta and Facebook, in a lot of those cases, you just put out a blog post and you're kind of done. But in this case, you know, there's a product running on the superhuman domain. It means, you know, Grammarly runs a certain way. And so we needed the whole company to execute. And so what we launched yesterday was, you know, a lot of it is somewhat invisible, but it's a new website. It's a new login system. It's new pricing. It's new everything.

31:07and we kind of had to do it with the whole company involved. And to do that, you kind of need the world to have some sense of what's coming. So, yeah, it was an interesting process, a little different than I think most renames. Rahul, you are the one sort of, you were acquired and Superhuman was your baby, including the name itself that you gave it. But I want to ask you, now that you're sort of operating within the larger Superhuman company, I guess Superhuman formerly Grammarly, now Superhuman, now it has the different product lines. Talk to me about how that's been different for you and also how this sort of reflects the changing strategy of the Superhuman company and the product suite more broadly.

31:51Well, I think one of the most interesting things about selling a company to a much larger company, Grammarly, is more than 10 times the size of the employee base, a significantly larger revenue base. as we turn into superhuman, of course, that continues to be the case, is you can simply do so much more simultaneously. So one of the most exciting things for the team was to be able to invest more in mail, to invest in our roadmap. There was this incredible moment multiple times during the process of acquisition when Shashir and I were comparing roadmap slides, vision slides, and they would even use some of the same terms.

32:27We would say things like, we want to build the AI native productivity suite of the future. And there was this moment where I was reflecting to the superhuman male team, where as a founder, you get to paint these grand pictures and big visions and say, we're going to do act one, then act two and act three. And as the numbers would probably imply, we're probably going to do act three after act two and act two after act one. But we're now in the position where we can pursue acts simultaneously. It's one of the most exciting things for a founder. Better together, as they say. That's always how it goes.

33:04Shishir, I want to come to you about this new product that you guys have launched, Superhuman Go, which to me, as best I can see, it's sort of like a browser assistant that helps you sort of work alongside all of your tabs. Is that the best way to describe it? I think I would think about it a little bit more like the evolution of Grammarly. So Supreme & Go is our new AI assistant product. It's a platform that brings you your own network of proactive and personal AI assistants directly embedded wherever you work. And the core insight really starts with the technology layer of Grammarly. It's the, what Grammarly really does is it brings proactive embedded AI to over a million different applications, websites, web applications, mobile applications, and so on.

33:48And we can read, interact, and edit directly in those surfaces. So I don't think, think about it. So it's not attached directly to a browser necessarily. That's one of the ways you can use it is attached to the browser. It also runs as a desktop application. I see. way as well. You can use it on your mobile phone. It shows up as a keyboard. What we've done, what the Grammarly team has done is over the years built this foundation, we refer to as the AI superhighway, that brings AI directly where everybody works right in their application. So it's a much broader play than just what I think a lot of browser add-ons are doing.

34:28And it means that if you think about Grammarly as the agent, the OG agent that works right alongside you. And it's almost like having your high school grammar teacher sitting on your side with a marker and able to help make every piece of writing better and better. We're now enabling a much broader set of agents to do that. So now if you're a salesperson, for example, you might have on one shoulder, you've got your grammar assistant helping you make sure you're writing accurately and getting your tone right. But you also have a sales agent that knows everything about your catalog and tells you you're talking about the wrong product or your calendar agent tells you, you can't meet tomorrow, you can meet the next day.

35:06All of these happen right in your surface. Why is it better than the other AI assistants that is out there? Are you betting that your models are better? What's the bet that you're making? I think in this market, the key bet is the difference between being interruptive and being immersive. If you think about the different AI providers out there, many of them are making a bet on changing people's behaviors. There's a set that are focused on what we call the chat experiences. You go to them, you chat with an AI bot. And sometimes that can be an amazing experience, but you have to remember to do it.

35:39There's others working on headless automations, go send off these AI agents to go do things on your behalf. Our job is to bring AI directly to where you work. Maybe to give you one fun stat that maybe will help contextualize this, we do over 100 billion LLM calls a week. and that works out to across about 40 million or so daily active users works out to a few thousand calls a day. So if you just think about that and say, if you're a really good ChatGPT user, for example, maybe use it a dozen times a day. In our case, we're running it every time you press a key, every time you open a new dock, open a new app, we're going to run dozens of AI calls and figure out how to help you where you're actually working.

36:19And it's just fundamentally different than having to go to AI. Rahul, I want to come to you. You operate an email platform. Now, one of the questions I've always had is the privacy element to all this. You obviously have AI reading your emails, being able to make sense of them. And this is not just a superhuman specific thing. I mean, all these email platforms that are out there, including Gmail, for that matter, all these browser assistants or browser tools. I mean, you have to be able to read the emails to take any sort of action related to them. How do you deal with privacy at Superhuman? And is the data stored?

36:54Or I mean, how does it even work? Yeah, great question. And as you can imagine, something that has been a top priority for us since day one. Otherwise, we simply wouldn't be where we are today. And so we make several promises to our users. For example, we promise and have stuck steadfast to this that when you use the software, the data that is your data isn't used by our third-party partners, whether that is OpenAI or whether it's Anthropic or any other model that we might be using in order to train their data. There's also a zero-day retention policy. So of course, I think everyone knows this, if you're going to use AI in or around your email, it will be transmitted to a model, but what we can promise is that it won't be retained and it won't be used to train data.

37:47We also promise that we're not going to sell or monetize user content. I think this is something that I know a lot of people here care about, but our product is the software. Our job is to make you brilliant at what you do, no matter what you're doing, whether it's collaborating with a colleague or whether you're producing content. It is not to sell your data. It is not to monetize your content. And I think that sets us apart from quite a lot of other very big companies. Great. Well, I want to thank you both for coming on. It's always a great conversation. That is Shishir Mehrotra and Rahul Vora from Superhuman and Superhuman Mail.

38:23Okay. The information wrapped up its 2025 WTF Summit in Napa Valley, California yesterday. Our founder and editor-in-chief, Jessica Lesson, shared some insights from the conference with our VC reporter, Natasha Mascarenas. Let's listen in on that. Jessica, you just stepped off the main stage at WTF. Literally. I try to kick off my shoes, Natasha. I'm not gonna lie. I hope no one can't see them. I was gonna ask, how do you feel first and foremost? Oh, I'm so excited. This is a real highlight for me. I mean, you know, like, getting to see your readers, talk to them, your whole, you know, the community we've built here.

39:01More than 300 amazing women. And I thought we had some fascinating conversations. You did a great job up on the stage. So it's a weird moment because it's like my notes of like stories to write, themes to take advantage of. It's all swirling around in my head. And you didn't even use notes for most of the interviews. I don't know how you cut it all straight. I keep them there in case. Someone challenged me on a stat. So I was good. I had my notes. But I remember when we were talking about this event eight months ago, you were talking about volatility being a key theme. Kicked off the night talking about volatility.

39:31now that we're after two days of conversations, what's the biggest takeaway on how leaders are navigating that, you know, the craziness of this moment? I don't know about you, Natasha, but people are optimistic. I mean, not in like a sort of diluted way. I mean, whether it was about AI safety, AI costs, the craziness of competing in venture, you know, it's certainly a challenging time, but I think there was an optimism that shined through. And I think it was, and everyone's going to blur, but I think it was Jenny Kohler who runs PWC's advisory practice. And someone asked her that question when I was talking to her.

40:10And she said something about like remembering the privilege and honor of this work or something like that. So I think there was an optimism through the volatility at the highest level, but also a lot of we don't know yet. We don't know yet. You know, I was sitting down with Rebecca Blumenstein, president of NBC News, and I've known Rebecca for probably decades now. And she's very optimistic about the future of the news business and video. But I felt like on many topics as well on what, you know, the change in Google search traffic means, what AI means, she was sort of honest. This is going to be the story of our professional lifetimes and we don't know the answer yet.

40:49So that was still there, too. It matches what I was getting from investors as well. Nina Ashadijan from Index was sort of said, you know, if you don't, if you tell me that you don't have a moat, you don't have defensibility, I'd rather you just say that versus telling me something that's not true. And if it's not true ARR, is it a version of ARR that I can at least understand versus you lying to me? That was a great answer. She was basically like, just tell me how you calculated the number and I'll decide. And I thought, and then Lucy Guo, who was one of the co-founders of Scale.ai has now started a creator economy company, Passes.

41:21I was pushing her this morning on, okay, you angel invest in startups and AI, like, how do you pick winners? And she seemed a little like, I think you're going to have some revenue in two years, you know, we'll figure it out. So I think she probably has a more rigorous rubric than that, but, you know, still wide ranging opinions. But I think investors, you know, they're wising up to the numbers at least, although I think it was on when our colleague Steph Palazzolo was interviewing Asha from Microsoft, I believe. We've had a lot of speakers today. And no, no, I'm sorry. It was on your panel. You asked about an open AI IPO.

42:03Oh, yes. I was like, you have to see if it's going to change the whole market. And what one of your panelists answers, I thought was a great answer, was like, we'd like to see these numbers because we kind of think they're better than people realize and that would buoy the whole ecosystem yeah there's a lot of big moments that people are looking forward to i also want to just quickly ask too did you sense a big difference between how big tech was talking about this we had leaders from microsoft youtube netflix versus the lucy guo's the campbell browns who announced you know her new startup on stage forum ai like did you get a sense of you know where you know if they differed on where volatility exists in this moment it's really interesting you know video to get to that video was such a huge theme.

42:43And I think on the, you know, talking to senior women at Netflix, YouTube, you know, it is a little bit like you get a sense that the ship is steady. Now, of course, I could have the opportunity to ask Mary Ellen Co., who runs YouTube's business, probably an hour after they announced a AI precipitated reorg that came with voluntary buyouts. So, but the message there is It's definitely sort of business as usual. I also think the start, yeah, I didn't feel like the big companies were going out of their way to do like the scrappy startup thing too, you know, to make, which we've seen in other eras.

43:22So, but I do think, I mean, it's such a wild moment in tech because there's so many unanswered questions, but things are doing well. Things are growing going up to the right. And so I think that is really, that was probably behind a little bit of the optimism. I know we didn't get into, in many cases, the sort of macroeconomic picture, which I think is the sort of thing that tips the scale at some point that changes the mood. But overall, I think both big and small companies felt like they had a lot of progress to report. Okay. Okay. I mean, last question, if I can put you on the spot. Yes. If you had to guess what the theme will be for next year.

44:02My goodness. For WTF. Is there a word that you're at least keeping in the back of your mind as you leave this event and start, I'm sure, planning next year's event? No, I did. I've been writing down speaker ideas all day long. How did it feel? Also because people come up to me and they're like, why isn't so-and-so here? Why isn't so-and-so here? You know, I have been thinking about next year because we're also getting to our year-end prediction time where we do a ton of coverage in the newsroom about that. And I'm kind of feeling like next year is going to remain high on this plateau of sort of on a lot of vectors, like on excitement, on asset prices, on all of these things.

44:39But but with that sort of macro question in the background. And so I think going into next year, it's going to be really important to start to maybe connect the dots more with some of the economists, market experts kind of outside. You know, through security, we talked a little bit about the geopolitical landscape and China and all of that. But I think that could come more center stage. But I think there are also going to be like three new AI shopping startups up on that stage who probably have billion-dollar seed rounds and, you know, will probably be using their products. And hopefully a few more IPOs to talk about as well.

45:18Yeah, although those are coming. Yeah. But overall, it's been great and really pleased. And you're right, looking forward to next year and really happy that we got so much coverage out on the website for people to kind of check out. And I hope it's really useful beyond the confines of this event. Yes. Well, I'm excited to get to the happy hour portion. Thank you so much, Jessica, for another great year. You're welcome. Thanks. Over on the main stage at WTF, Jessica sat down with Lucy Guo, founder and CEO of Passes, a platform that helps creators monetize their brand and keep roughly 90 % of their earnings.

45:53Here is what she had to say about how creators are using AI. creators are actually like very against ai tools i'd say at the moment um but i think it's obvious that ai helps us scale our time and i think the future and maybe like the next five years is creators are going to license out their likeness um so instead of having to like spend a week working with their brand flying out having to you know book a hotel get like you know makeup artists a videographer editors etc um they'll just license out their likeness so they can focus their time on actually creating content and growing their audience.

46:25Obviously, they can't do this with every brand because they'll dilute their brand, but I think that it will help them scale their time, and brands will end up paying less money per creator, and creators will make more money by being able to work with more people because they are scaling their time. But right now, they're just not really trusting of anything related to that because they're... Say more about that. What's the fear? So I think there's this idea that tech companies are big and bad and selfish. and that we will end up using their data and they won't earn the cut that they deserve to earn.

46:59I also think creators are in a unique place where like, you know, in tech, we love AI because it makes us more productive. But for them, they're scared about their likeness being used in ways that will tarnish their brands. And we're already seeing, you know, people make deep fakes of creators saying inappropriate things. And they're worried that this will happen if we are able to collect all their data. Got it. So what needs to happen? I mean, I guess Paris AI, which we heard a little bit about last night, might be a version of this. But what needs to happen to enable that, you know, hey brand, you can use my likeness.

47:35I'm not coming. Probably don't get paid as much as if you showed up. But are we there yet? I think we need some open-minded creators and we need one creator to make a lot of money because I think creators are very driven by wealth. I think there's a fascination with tech because of that, where, you know, like after Kylie made her lipstick brand, every creator wanted to start a company. And we're seeing that, right? I think the first creator that's super open-minded, that's able to license out their likeness, and they start making like eight, nine figures, other creators will follow suit. veteran news anchor and former meta vp campbell brown recently raised three million dollars in seed funding to co-found forum ai a new company focused on fact-checking ai services like chat bots with a special emphasis on identifying bias around hot button issues natasha spoke with brown exclusively at the wtf summit let's listen in on that conversation hi everyone i'm natasha we are at the WTF Summit.

48:36We've had a great day full of panels and discussions, which we're really excited to get into. For now, I have an exclusive interview with Campbell Brown, the founder and CEO of Forum AI. Campbell, thanks for being on TITV. Love being here. Thank you for having me. So tell me, let's jump right into it. What is Forum AI and why launch it today? So this is a startup that I've been quietly working on with a co-founder for about eight months. We've just come out of stealth and are starting to tell our story. And this came out of, I think, me watching teenagers, I have two teenagers, use chatbots for their primary source of information and recognizing that the information they're getting is sometimes pretty bad.

49:20And there's a real need for, especially around complicated subjects, there's a real need for more depth, more clarity, more accuracy, obviously. And so we're focused on trying to do evaluations of the outputs of all the big models on high stakes topics. So think politics, conflict, geopolitics, and even mental health, which, you know, if you have a teenager using these bots to ask mental health questions, the tone is really important in terms of getting it right. So So we're first evaluating the outputs of the models. And we do that with AI that we've trained with a network of experts. And the experts are people who range the spectrum of politics from right to left and everything in between.

50:11We've partnered with a lot of Washington think tanks like the Atlantic Council, Manhattan Institute. We've partnered with Cleveland Clinic and Mount Sinai and the Stanford Institute for Human-Centered AI. So a lot of academics who can work with us, too, who can evaluate what's missing. When you get an output, is there a perspective missing? Is there context missing? And then when we find out what this needs in order to be a broader, more balanced perspective, we have experts in the network who can create content and provide some of those answers to really tough questions. So people like Fareed Zakaria and Neil Ferguson talking about geopolitical issues.

50:51they both strongly disagree, but they're both intellectually honest. And so, you know, if they were to look at a piece of content, they might say, yeah, this needs the, you know, it has my perspective, but it needs the other side here. And so it's a little bit like peer review. Okay. Yeah. And walk me through that process a little bit more because, you know, when I think about that, I think, is there a future where you're going to have a little check mark at the bottom of answers saying approved by Forum AI? Is it more something that's going to be completely invisible and before we see OpenAI or Anthropic perplexity announce our next product, you'll have been behind the scenes?

51:26How visibly does this show up? So right now, we're very much behind the scenes, but that's a really interesting question because you can see where this is kind of headed already in terms of the regulatory environment. The EU is sort of mandating, not sort of, they are mandating bias mitigation. There's lots of talk in Congress on Capitol Hill around how to address bias in the models. But it's not just bias. It's also just getting this, you know, the right tone, the context, the broader picture. Everything is not left and right. And, you know, that's often how it's presented. And I think especially really complicated issues coming from these models.

52:10It's a lot more nuanced than that. And what we want to try to do is bring this really diverse expert network into the process so they can bring some of that nuance to the answers. Is there a certain sort of category of question that you feel like Form AI specifically excels at? And, you know, an area that you see yourselves needing to get better on and you see it, you know, being better in the future. I would love to sort of know the here and the future. So we started with what we thought were going to be the most difficult. Okay, okay. But it was also the area that I have the most experience in, which is politics and world affairs.

52:48Both that, you know, you look at all these data labeling companies that are very successful in Silicon Valley. They stay away from those issues because they're too controversial. And who wants to touch those? And even one of my investors even sort of jokingly said, why are you breaking into jail? Like, why do you want to do this? It's like, because this is important. These are really consequential issues. and we need to make sure that we're getting these right. So we started building the network of people who have a lot of experience in politics, and then from there, world affairs and geopolitics, and partnering with a lot of the think tanks, which got us the, you know, the breadth and depth that we needed.

53:26And then through our conversations with the big AI companies, mental health kept coming up, is that a lot of teenagers are using these tools to ask difficult questions, especially around mental health. And, you know, you can imagine first the liability issues with that, but also how important the nuance and the tone of that is. It can't just be a step-by-step medical response. It has to really be able to speak to someone who may be in crisis. And so we started, we're working now with Cleveland Clinic in Mount Sinai in trying to address some of those questions. first in the evaluation piece and then again, providing new content, you know, as needed to try to fill some of those gaps.

54:13And yeah, just to double click on the Cleveland Clinic example, because that's super interesting. Are they approaching you and saying, we use these tools internally and we want to get better answers for our doctors and patients? Or yeah, who is, you know, where is their pressure point? Well, so we reached out to them. So all our partners, we reached out to them. And for them, it's very much part of their mission to make sure accurate, good information is out there. And I think most of the think tanks feel the same way. They're generally nonprofits, and they want people to get this right. And so they were excited by the opportunity because it gave their people, A, a chance to show their stuff.

54:54They had been creating lots of content intended to try to achieve that goal. And I think a lot of us see that move toward whether it's ChatGPT or Claude or MetAI or Gemini, whatever you use, is sort of being the go-to for people in the future and wanting to make sure that they're present and that there's attribution there for these more reputable sources and that they're getting highlighted. And, you know, one of the big problems the LLMs have, it's not necessarily a shortage of information. It's just more, how do you find the signal and the noise? And that's what we're trying to help them do is like looking at the evals and doing like really nuanced labeling that can give them instructions for how to get to a better answer.

55:40Yeah. I mean, and it's not a topic that you are at all a stranger to. Obviously, beyond being a renowned journalist, you were formerly at Meta as the head of news and the VP of media partnerships. From that experience, you know, is there one core learning that you're bringing with you that you can kind of speak to about the way you should approach that conversation with the companies? Yes, very much so. I mean, we tried so many things at Meta. Some worked, some didn't work. But, you know, from our fact checking program to starting Facebook news and working more closely with news organizations and working through these issues and trying to get the balance right was really hard.

56:21And no one ever thought we did. You know, I don't know if we did, but I'm not ready to stop trying. Yeah, yeah. So that was partly, you know, inspiration for it. And then being at Meta just as AI started to really explode. And, you know, the company pivoting its focus to AI and it becoming really clear to me that where this was headed in terms of, you know, being a source of information. and seeing both opportunity in that and, A, it being even more important that we get ahead of some of the problems that we had at Meta that we were trying to kind of clean up. Yeah. I mean, and when we see the kinds of partnerships that already exist between tech companies, AI companies, and media businesses today, I think about the licensing deals.

57:09I think about those content partnerships. Is this meant to be complementary to that or competitive to that? Complementary. Okay. I mean, I very much hope that this continues bringing news partnerships to the table. You want to see more of those? Oh, yeah, yeah, yeah. Okay. But, I mean, partly because news is struggling to find a business model. And I think those partnerships are pretty critical. But news organizations have a lot of bias. And how do you balance the content that you're getting from news organizations? And who has the expertise to look at that and ensure that when you're generating some AI output, that that balance is there?

57:50And, you know, look, I have no control over where these chatbots are ultimately going to go. My hope is that it is not the path of the filter bubble where, you know, whatever your AI of choice is gets to know you so well and only presents you with that one narrow perspective that it knows you want to hear. You know, I'm not that way. I want to see lots of different points of view. I don't want my children to be that way. I want them to learn from other people who disagree with them. So I want to at least try to have that present and accessible. And for the most part, it seems like the model companies are interested in having that too, which is just the broad, like I said, the depth and breadth of content that's out there.

58:39No, it's a good thing to hear. And obviously, you've raised some money. You've hired a team. I'm so excited to check in with you on this, hopefully a year from now, and hear more stories like this. But for now, thank you so much for joining us on TNTV Campbell. Thank you for having me. That does it for today's show. A reminder, we are on this stream Monday through Friday at 10 a.m. Pacific, 1 p.m. Eastern. I want to thank Amazon Web Services, who is our presenting sponsor for this production. And I want to thank you for tuning in. We really do appreciate your viewership. I am already excited for our next show tomorrow.

59:10Have a great rest of your Thursday. Bye-bye for now.