Pod Insights Big Technology Podcast The Sora 2 Breakthrough, Death Of The Creator Economy?, Apple’s SmartGlasses Roadmap About this episode Max Zeff, Sr. AI reporter at Techcrunch, joins for our weekly discussion of the latest tech news. We cover: 1) OpenAI introduces Sora 2 2) Will these AI video feeds catch on? 3) Is Sora 2 an important technological advance above all? 4) Why Meta is nervous about OpenAI's momentum in social 5) OpenAI employees have mixed feelings about Sora 6) Why AI video feeds may be the end of the creator economy. 7) Why they may not be 8) Meta will use AI chats to train its ad models 9) Meta's AI research moves more toward product development 10) Anthropic launches its Sonnet 4.5 model 11) Apple prioritizes smart glasses 12) Max tries out the Friend pendant
---
Enjoying Big Technology Podcast? Please rate us five stars ⭐⭐⭐⭐⭐ in your podcast app of choice.
Want a discount for Big Technology on Substack + Discord? Here’s 25% off for the first year: https://www.bigtechnology.com/subscribe?coupon=0843016b
Three Faces Of Generative AI: https://www.bigtechnology.com/p/the-three-faces-of-generative-ai
Questions? Feedback? Write to: [email protected]
Learn more about your ad choices. Visit megaphone.fm/adchoices
Transcript 10,250 words · 3 speakers
0:00 Speaker 00 The gates of content hell are open as OpenAI's Sora 2 debuts with ultra-realistic AI slop, and Apple has elevated its smart glasses plan, and they could be coming soon. That's coming up on a Big Technology Podcast Friday edition with TechCrunch Senior AI Reporter Max Zeph, right after this. This episode is brought to you by Genesis. What does it actually take to put a Gentic AI to work across an entire enterprise? I recently spoke with Genesis Chairman and CEO Tony Bates about why orchestration is emerging as a competitive advantage in customer experience and what companies need to get right in the AI era. Then, I sat down with Adam Mitchell, Head of Enterprise Business Solutions at Voya
0:44 Speaker 00 Financial, to go inside a large-scale agentic CX transformation, including the business decisions, trade-offs, and lessons along the way. You can watch both conversations now on my YouTube channel today. It's never been more challenging to get frontier compute. How do you know where to look, who to trust, or how much to pay? Akash is an open cloud that provides easy access, flexibility, and peace of mind. That's how companies like Razer were able to cut their inference cost by 15 times. Visit akash.network/alex for a free consultation. That's akash.network/alex for a free consultation. Welcome to Big Technology Podcast Friday edition, where we break down the news with
1:28 Speaker 00 our traditional, cool-headed and nuanced approach. Today, we have TechCrunch Senior AI Reporter Max Zeph joining us to talk about everything AI slop. There's so much slop, but maybe it turns into something. We'll talk about how it might challenge the creator economy. We'll also talk about Apple's smart glasses and the fact that it's speeding it up in the roadmap, and plenty more. Max, great to see you. Welcome to the show.
1:52 Speaker 01 Great to be here. Longtime listener.
1:56 Speaker 00 Definitely. Well, we are, we always talk about your stories on the show. Here is a hint for listeners and viewers. When you hear me cite a TechCrunch story, it's almost always Max's. And Max, we're going to get into yours because you have some new reporting on the state of OpenAI after this Sora 2 launch and how some of the employees there are trying to figure out what it means for the company. But let's start off talking about Sora 2. For those who don't know, Sora 2 is OpenAI's new video generator, which they've now included in an app called Sora, which is basically meant to rival TikTok and YouTube. This is from the Wall Street Journal.
2:39 Speaker 00 "OpenAI is squaring up to TikTok, Google's YouTube and Meta platforms with a new social media app for its AI video generator that allows users to create high-definition video clips with audio from text prompts. Users can upload chore clips of themselves and insert them into Sora-generated worlds describing the idea, style and scene they want to see. They can also connect with other users watching and commenting on their content." Max, just first question for you. We've, we're now a couple of days into this Sora madness. It's the number one app on the App Store. How big do you expect this to get?
3:15 Speaker 01 Yeah, I mean, I could see this taking off quite significantly. I think in the last year, we've really seen AI-generated images and videos really take over regular social feeds. And I think the idea here from OpenAI is let's capture some of that growth on one of our own platforms. And while it sounds terrible and, you know, at first it kind of makes you cringe at the idea of a completely AI-generated feed, it's been surprisingly compelling, I think, to a lot of people. As, as you can see by the growth on the App Store and just my own personal experience, you go on there and it's like, this is terrible. And I, it's like you can't take your eyes away. It's like watching a car crash on the side of the road.
4:01 Speaker 00 Yeah, I think it's worth pausing just to talk about how realistic these videos are. I watched one of Sam Altman on the release day, running around OpenAI headquarters with a bullhorn, telling them to get to the live stream. In the first few seconds, this was like one of my first Sora2 videos that I had seen. I was just like, that's real. And I was like, wait a second, that's not real. And then I really got mad at myself this week because a friend of mine posted a video of Jake Paul coming to their book signing. And, you know, my friend signing a book and handing it to a very happy Jake Paul. And these videos are so good that your brain doesn't process the fact that they're AI-slap at first.
4:43 Speaker 00 And I was like, you know, when you see this stuff, you go through these machinations in your mind of like, how did that happen? And I didn't know that he had a book. And then I was like, oh, my God, that's Sora. Like, it's crazy.
4:56 Speaker 01 Yeah, there was a really interesting take from a former OpenAI employee, Miles Brundage, who kind of came out and said that, you know, we're getting into this era of AI-generated video feeds. And it's worth noting that not all AI-slap is created equal, which I thought was a surprisingly thoughtful take about something that seems at its face really kind of scary. But I think he's right. I mean, I think as we saw with Meta's Vibes app last week, that's Meta's TikTok competitor that's full of AI-generated videos. And that product doesn't let you make videos of yourself or your friends. And I think the reaction was people were like, well, what is this? What do I do with this?
5:41 Speaker 01 Whereas with OpenAI's product, it's immediately become really compelling because people can see themselves in it, they can see other people they know. And, and that seems to be a really key detail here that is a big differentiator.
5:57 Speaker 00 You get the sense that two things happened. First of all, Meta thought that Meta must have gotten word that OpenAI was going to release this and then rushed out vibes. And I think the other broader picture is people started to question why Mark Zuckerberg was sinking so many billions of dollars into AI research and recently going on this hiring spree of AI researchers. I think it's worth noting that a lot of the researchers that he poached were working on multimedia formats, audio, video, images. And you see, I think this week is the first time it's really come into focus, that this could actually challenge the standard
6:40 Speaker 00 human-created content feeds. And that's why it's like a nine-alarm fire inside Meta HQ right now.
6:48 Speaker 01 Yeah, I mean, I think that this is the change that people have been talking about for a long time. I think it was hard to see for a while why a company like OpenAI was threatening to Meta. And I think this week, you see why. I mean, I think that this is very clear evidence that there is something here. I don't think that, you know, this product is necessarily the final version of it. I'm not sure if it has going to have a great impact on society at its, you know, at the net. But I do think that this is compelling. And a lot of people spent a lot more time on the Sora app in the last week than they did on Threads or Blue Sky. And there's something to be said for that alone.
7:31 Speaker 00 We had a celebration for five years of Big Technology Podcast this week in collaboration with Human Ventures. They hosted here in New York. And we gathered a bunch of people around and we were talking about this and what the future is going to look like. And Rajan Roy, who's usually here on Fridays, told the story that he tells on the show so often about how his Facebook feed had become just a feed of AI slop and people commenting on it. And I think there had been some questions about why Mark Zuckerberg and Facebook were allowing the feed to fill with like Shrimp Jesus. And again, this week is like bringing it into focus that that it is something that is going to be, maybe already is, a very engaging
8:15 Speaker 00 content format. And if you are the engagement app, which is, or the engagement company, which Meta is, You can't really let another company run away with it. And maybe they already are seeing OpenAI run away with it.
8:29 Speaker 01 Yeah, I do have real concerns about what the impact is on people who use social media and society for something like this. I think that there's a sense to me of, you know, these apps, you really, at their core, are supposed to be about connecting you to other people. And I think we've seen all the ways in the last, you know, two decades of how that's not really the case and how they really are optimized to keep you on the platform. I think there's a big question for OpenAI here of what they really choose to optimize for in this experience. I think we've seen some good players emerge in the social media space like a Blue Sky
9:09 Speaker 01 that doesn't seem to be focused on getting people to, you know, just spend as much time as possible on the platform all the time. That said, I think it limits their growth and I think OpenAI is a very ambitious company. I think there's a real chance here that, you know, they are pushed by some of the incentives that social media companies have been pushed by for, you know, the last 20 years.
9:34 Speaker 00 But it is different and I think it's worth taking this moment to just pause and talk a little bit about the underlying technology and how impressive it is and why, to me at least, this is not just OpenAI creating a frivolous, you know, vapid social media app, but how this might be the next point along a line of research that needs to get done to improve this technology. So there's been so much talk about AI's need to understand the world and people have talked about world models, the understanding of physics, understanding of what happens beyond the LLM.
10:12 Speaker 00 And there was an interesting point here that a user on Twitter or X made that sort of puts OpenAI's lofty ambitions into contrast with what it ship. Sam Altman, two weeks ago, we need $7 trillion and 10 gigawatts to cure cancer. Sam Altman, today, we are launching AI slot videos marketed as personalized ads. Sam Altman chimed in. He said, I get the vibe here, but we do mostly need the capital to build AI that can do science and for sure, we are focused on AGI and with almost half, sorry, with almost all of our research effort. It's also nice to show people cool new tech products along the way, make them smile
10:53 Speaker 00 and hopefully make some money, given all the compute that we need. When we launched ChatGPT, there was a lot of who needs this and where is AGI? Reality is nuanced when it comes to optimal trajectories for a company. Just this line about the launch of ChatGPT, who needs this and where is AGI? I just think it's important. I mean, this is sort of the key, the key part of Sam's statement. And I'm not, not here to be like Baghdad Bob for, for Sam. But I will say that when you get this stuff, starting to understand physics, you're just a step closer to, I think, the end goals. And ChatGPT, of course, like seemed like a little toy at first, but obviously has emerged
11:37 Speaker 00 into something more. And I get this, the criticism, but I tend to side with Sam on this one. What do you think, Max?
11:45 Speaker 01 Yeah, I think it's, I think it's a tough call. I think I agree with Sam in some ways that he's right that ChatGPT, it wasn't really understood at first how this really was going to be a vehicle for a lot of people to use AI and a lot of people to actually get the benefits of AI. And I think we've completely seen that play out in the last few years. I think it's a little less clear to me how Sora gives, gives people the benefits of AI world models in the same way. I totally agree that the underlying technology is really impressive. I mean, when the first Sora came out, I think last year, it blew people away just with how realistic the videos were, with how accurate it could get certain details.
12:29 Speaker 01 And this one is objectively better. And I think it's impossible to argue that there hasn't been a great leap in performance of the AI video models in the last year or two.
12:41 Speaker 01 But ChatGPT at its core is about helping people. That is, I do actually take OpenAI at its face when it says that it optimizes ChatGPT to be helpful and not to just keep users in, you know, these engagement loops. I do think that I go to ChatGPT all the time to get an answer. I get it and then I get off the app. I think with Sora, it does feel to me like it is, I mean, he said it pretty plainly. It's about entertaining people. It's about, you know, it's supposed to be fun. And I think that these are very different products. And while I agree that there's something to be said for OpenAI's ability to make products,
13:21 Speaker 01 to fund, you know, its AGI efforts, I think that it's worth noting that he's venturing into a very different space than he's been in before.
13:30 Speaker 00 Okay. Let me talk a little bit about what grounded my comments and then we're going to actually talk a little bit about your reporting that has OpenAI employees basically expressing similar concerns. To me, the thing that I've heard from AI researchers is you really want AI to be able to predict what's going to happen in the world, because when it can predict, then it can plan. And predicting and planning are core parts of intelligence. Now, the thing about chat is, or text is, it's somewhat easier to predict the next word than it is to, let's say, and this is a Yann LeCun example, hold up a pen and drop it and predict where it's going to go because there's seemingly infinite possibilities.
14:14 Speaker 00 So when this technology gets better at understanding like a person, what a person walking on a sidewalk does, then that's a step forward in its ability to predict and then plan. And so the underlying, maybe not the social app, but the underlying advances that they're working on here to me seem like an important step forward. But yeah, I could be getting over my skis here. So what's your thought on that, Max?
14:38 Speaker 01 I think you're right. I mean, I agree that world models are an important next step for AI. There was just a viral podcast where a famous AI researcher, Richard Sutton, went on the Dwar Keshe Patel podcast. And there was this really fascinating argument about how LLMs today, they don't have a model about the world. I mean, they can predict the next word, but they can't predict the state of the world. And they can't predict, they can't update their mental model about the world based on interacting with it. And I thought it was fascinating conversation. I would encourage people to check it out.
15:17 Speaker 01 But it was really, I think, you know, this is something that Jan LeCun has been, like you mentioned, talking about for a long time, which is that world models are really a key to kind of a greater intelligence and greater ability to predict the world. So I do think that it's important for OpenAI and these AI labs to advance the world models. But yeah, I think the application is really the key part here.
15:44 Speaker 00 Okay.
15:46 Speaker 00 So briefly, so your perspective is the advances in video that we're seeing might advance world models or does advance this concept of a world model, but just the feed is where you start to feel a little queasy?
16:03 Speaker 01 Yes, 100%. I think it is like, you know, you can take a good technology and use it for the wrong thing. I think is my feeling. And I just wonder if this is that.
16:13 Speaker 00 Okay.
16:15 Speaker 00 Yeah. And you had some reporting that shows that even people within OpenAI are concerned there. And I think this is a great place to focus and stop for a moment. Here's your story. Several current and former OpenAI researchers are speaking out over the company's first foray into social media. "AI-based feeds are scary," said OpenAI pre-training researcher John Hallman in a post on X. "We're going to do our best to make sure AI helps and does not hurt humanity." I mean, where have we heard that before? You had actually multiple current OpenAI employees being like, "I don't know about this. That doesn't happen too often in a major launch." So talk about the significance of that.
16:55 Speaker 01 Yeah, I think the level of kind of concern expressed by a lot of OpenAI researchers at the launch of this product felt very unique to me in a way that you don't usually see. Usually, when an AI company launches a product, everyone is very gung ho about it. Everyone is really excited, and I think this time just felt a little different. I think a lot of researchers at OpenAI were very aware that their company was not just launching a new product. It's not a new feature in ChatGPT. This is a social media app. This is uncharted territory for OpenAI. And I think that raised a lot of concerns for people. I mean, something I've heard over and over from AI executives, and I'm sure you've heard
17:39 Speaker 01 the same, Alex, is just they all talk about how we don't want to repeat the mistakes of the social media era. And I believe a lot of the researchers working at these labs don't want to do that. I think a lot of them grew up in an era where people were addicted to social media. And I think a lot of them don't want to create the same kind of dynamic with the products that they built. But I do think that just the incentives of running a social media platform are sometimes going to be very intoxicating and strong. And I don't think Meta exactly set out to build addictive platform, but that's what
18:23 Speaker 01 you do when you need to make the user count grow up and you need to make the ad revenue grow up. So I think that that's what I heard from a lot of employees, just like, we have to be really careful here.
18:36 Speaker 00 Right. And I think it's, we should note, head of product at Anthropic, Instagram founder, head of product at OpenAI, former head of product at Instagram, person who's running what all consumer applications at OpenAI, is also someone who's come from Facebook. It's interesting to me to see these two cultures merge, which is like the typical Silicon Valley growth culture and this newer research-oriented AI culture. And effectively, we can see the conflicts happening within the companies as they go.
19:19 Speaker 01 Yeah, I mean, this is something that's come up with OpenAI time and time again, where there's this tension at the company where it is the world's fastest growing consumer business. And it is also an AI research lab where a lot of the people joined to chase this lofty mission of developing AGI that benefits humanity and make sure that all of that goes safely and goes well. And I think those things have constantly been just kind of pushing at each other for the last several years. I mean, when Sam Altman was fired a few years ago, that was, you know, kind of a key consideration there.
19:59 Speaker 01 And I think that what I heard from a former employee that was really interesting is they said that, you know, OpenAI really, the consumer business funds the mission. I mean, that's how a lot of people see it, where it's like, you know, building ChatGPT into this massive product makes billions of dollars that can go towards AI research that otherwise wouldn't. And I think that's true, you know, to build AGI, you probably need to be an enormous company. But I think there's a real question of, at what point does OpenAI say no to a money-making endeavor? At what point does OpenAI reject the idea of, you know, adding more users to its platform
20:43 Speaker 01 for the sake of its mission? I think that it is, my take about the store app is I don't think that OpenAI is setting itself up for success with this because they are opening themselves up to, you know, the kind of incentives of a social media app.
21:01 Speaker 00 Right, yeah, you asked that question in your story. At what point does OpenAI's consumer business overtake its nonprofit mission? I think it's already there, so. But anyway, we can leave that open since you left it open as a question in your story. I won't, you know, put you on the spot on that one. But yeah, it's a valid, valid concern. And, you know, I think we can both agree that this is not just a video generation technology, it's a social app. And it does sort of then put forth this really interesting question, which is what's going to happen to the creator economy. I'm writing about this in Big Technology this week. It'll probably come out right before our, our episode comes out, but you have this,
21:45 Speaker 00 you know, creator economy is kind of a misnomer. There's only a small amount of people that are able to work full time creating content on the internet. That being said, some of the people who have watched this have said basically like this puts creators right in the crosshairs of AI. And in fact, I had a, I guest hosted the Twit podcast over the weekend and I had a great quote from one of our panelists, Ari Paparo of Architecture Media. He said, "The creators are all high in the hog right now, but I would be thinking they're one of the first things that could be totally annihilated by AI." I think it's worth taking a minute to talk about this because you have this new technology and new app. It's obviously very engaging, number one in the app store, but instead of the
22:26 Speaker 00 painstaking process of creating content online, you could just prompt it. So what do you think this means for the creator economy, Max?
22:37 Speaker 01 Yeah, it's really an interesting question. I mean, I think that people said some version of this for writers when ChatGPT came out that, you know, why would I hire a freelance journalist to write something if I could just ask ChatGPT to write it? And I mean, certainly there are countless websites out there with articles written by ChatGPT now. So to some extent, that's proven to be true. And yet, Alex, we, we still can make a living doing what we do. So I think what have we learned from the text version of this story that's already played out is that people value the human touch and the human taste and judgment, and they value kind of a professional's opinion, you know, on certain topics. That
23:21 Speaker 01 has become, actually, I think, more valuable in the AI era. And I think that was something that a lot of people missed when ChatGPT first came out, which is that it actually made really high quality human content more valuable. I wouldn't be surprised if a similar thing happens with the creator economy. Like, I think that- We will see just more content in general as a result of these AI video models and Sora. And I think that it probably does make it harder for content creators to stand out. It probably, advertisers are looking at these AI video models and going, huh, that's pretty interesting. Should I maybe put some of my revenue or my ad spend towards that instead?
24:06 Speaker 01 But I think that there's creators out there who have a real relationship with their audiences. And the audiences care about what they have to say in the same way that people care about what journalists have to say. Their favorite ones at least. And I think that that's kind of how I expect this to go in all of the industries that AI will touch. But I don't know, Alex, I'm curious to hear what you found talking to some people in the space.
24:33 Speaker 00 Yes, so I actually have this like list that is basically my outline for my story of all the reasons why the creator economy won't go away. And I think you've touched on a couple of them. It will change though, in some ways. But I would say like the first thing that I put on the list, which is somewhat corny and cliche is just the word authenticity. And I think that when people are saying the creator economy is going to go away, they might be thinking of these like adventure videos on YouTube or like some others, I don't know. But I think they are missing the fact that the reason why the creator economy or whatever it is emerged is because people wanted a more authentic alternative to, let's say, the mass media.
25:17 Speaker 00 And so they were able to have this connection in some ways with online creators. And that powered the creator economy. So I think that you cannot replace that authenticity with synthetic media. You just can't. It's impossible to create. I think I could be proven wrong over time, but I think that's impossible to create. The other side of it is. That that basically, the reason why you would tune into someone is for new insights, or ideas, or information. And even if you could generate amazing videos, you're just recycling what exists. And you lose that, you know,
26:02 Speaker 00 potential to gain those new insights and information, or even, you know, unique new entertainment, I think. So, so to me, that is, that is two of the main reasons why I think the creator economy is not going away. Now, I do think it's going to change, because there are some things that are just going to be commoditized.
26:26 Speaker 00 And I'll just say it, I think that if you're a creator that's making money on good looks, you might be in trouble. You might now see an AI, there already are AI versions of people who are doing like thirst trap content, and they're gonna, they're gonna commoditize that. So, also, this just sort of like, there's a lot of like amorphous mid-content out there, that maybe if you filled a niche, you could make a living, you probably won't be able to do that anymore. So that's kind of my read on it.
27:00 Speaker 01 I mean, I think that's a fair point. I mean, I think that that's kind of the flip side of what I was just talking about with the journalism conversation of, you know, I think that a lot of the websites that made a business on aggregation of just taking other reporting and kind of repackaging it, without adding some unique analysis, or, you know, new reporting to it, they, that largely can be done by AI. And I think that that's what, you know, that kind of, I'm not trying to say that's lower quality. I mean, I've done that fair amount of that in my career, but like, I think that that is Part of journalism that is kind of going away in the AI era.
27:42 Speaker 01 And, you know, I think I could totally see a similar thing happening in the creator economy. I will say, I'm curious what you think about, kind of, you know, obviously, I mean, the creator economy is not held up by TikTok and Instagram paying out these creators. It's held up by like AG1, like these like, you know, massive advertisers that like will, you know, just like dole out money to the creators. And I'm not sure if they're, you know, going to want to put their ads in Sora or something, you know, like, like, I think that that kind of source of revenue seems pretty stable if you're a creator, but I'm curious what you think.
28:18 Speaker 00 Oh, I disagree on that one. I think advertisers are very simple animals. If there's attention, they will pay for people's attention. So even if it's a feat of AI slop, if OpenAI can somehow win millions and millions of users over, then AG1 will be right there and work out some deal to sell their very healthy supplements and all the other things. So AG1, if you're listening, big technology's open for business. So we'll be human until we're not. So let's talk. But that's basically my take on that front.
28:51 Speaker 01 That, no, that makes sense. I mean, that you totally could be right. And I mean, Mark Zuckerberg has his idea about, you know, automating all AI, the whole ad process with AI that, you know, it's been talked about, you know, a few weeks ago. And, you know, I think a lot of advertisers' ears maybe perked up and maybe they cringed a little bit at that thought. But, um, yeah, I, I guess we'll have to see. I do think that there is some value in like, your attention is deeper when you're watching a creator that you love. So maybe that will, you know, prove to be resilient in this era. But I guess we'll see.
29:28 Speaker 00 Yeah, I don't want to be the gloom and doom on the creator economy ad thing, but.
29:33 Speaker 00 Ultimately, like they, the advertisers back it out like they're, they're smart. They will see like what's leading to sales, what's not, and then just keep investing more in that. Maybe, who knows, the meta can tune feeds in ways or OpenAI will tune feeds in ways that will make people more susceptible to this type of advertising. It's, it's, it's a, it's an interesting world. And by the way, speaking of automation, I want to bring this point up before we move on from this. There's a concept in artificial intelligence called reinforcement learning where like you give the AI a goal and it sort of makes a bunch of attempts. And then, you know, basically tries to accomplish that goal and then realizes what to optimize for.
30:17 Speaker 00 I, I had a thought that could we end up seeing, so, so while I think the creator economy is going to stay, do we end up seeing AI creators emerge where companies basically give bots this goal of like being mega influencers, allow them to create lots of accounts and then just sort of reinforce based off of the comments and likes and shares and follows. And so we end up seeing an AI creator force that will crush any human attempt to be, to be better than it just because of its ability to basically create unlimitedly and in an unlimited fashion and optimize that way.
30:57 Speaker 01 Oh God, I think that's going to be Meta's next AI benchmark and their next model releases, you know, how well can our AI grow its platform on social media? It's honestly a terrifying kind of example, but I could totally see that happening. I mean, RL has proved really good at doing specific tasks. Social media is a place where there is so much data coming in. It honestly is probably a pretty easy place to do something like that. But wow, yeah.
31:27 Speaker 00 RL for social media influence. Yeah. Yeah, it's great at playing games. Social media is a game. The retweets are the are the are the reshares. That's the points. So, so maybe we see it. All right. In the second half, we're going to get into how Meta is going to use our so our, our chat data or our engagement with its L L Ms for advertising. And then we're going to touch on the launch of Anthropic's Sonnet 4.5 model. And, and we will also cover the fact that Apple has seen what Meta is doing with its AI glasses and decided to accelerate its own smart glasses roadmap.
32:09 Speaker 00 So, you know, as I kicked a break, I'm just thinking to myself, wow, like, you might want to bash Meta, but it's in the conversation all across the board, whether it's hardware, advertising or content feeds. Something will work for that company. It always does. All right, we'll be back right after this. This episode is brought to you by App Point. Everyone's racing to roll out AI right now. Co-pilots, chatbots, agents doing real work. But here's the part nobody loves talking about. All that AI runs on your data. And most teams have no single way to see it, secure it and prove it's under control. That's exactly what App Point does. For 25 years, they've been the trusted layer beneath the world's most demanding
32:52 Speaker 00 data, now extended across your entire AI estate. Your data, your cloud, and the agents acting on your behalf. It's how more than 28,000 organizations deploy AI with confidence. So innovation scales without scaling risk. App Point, the unifying trust layer for AI. Head to apppoint.com to see how enterprises deploy AI with confidence. Learn more at avpt.co/bigtechnologypodcast. That's avpt.co/bigtechnologypodcast.
33:28 Speaker 00 Violent threats against executives are growing at an alarming rate. When a company experiences backlash, executives are the first people blamed. When their address and other personal details are sitting online, this backlash can lead right to their front door. The team at Ironwell knows this better than anyone. They've protected some of the most targeted executives and individuals on the planet for almost two decades. As a wide-glove enterprise security service, they protect your people with continuous personal data removal, proactive prevention tools, and human-led emergency support. So when someone goes looking for your executives, Ironwell ensures they hit a dead end. So here's what to do. Go to ironwall.com/bigtechnology, fill in the quick form, and request your free risk assessment. Their team will show you just how exposed your executives are
34:13 Speaker 00 and how to lock it down before a threat reaches their front door. That's ironwall.com/bigtechnology. Learn how to stop online threats before they become real-world attacks. The link's in the show notes.
34:28 Speaker 00 And we're back here on Big Technology Podcast Friday Edition with special guest Max Zeph. He is the Senior AI Reporter at TechCrunch. Max, you had a story this week talking about how Meta is going to use our AI chat data to target ads. And this has been something we've long worried about here on the show. Not the fact that it's going to happen, but the fashion with which it could be done. Because people are pouring their hearts out to these bots. They're using them as therapists and lovers. And is there an expectation? It means a tech product. Is there expectation of privacy? I don't know. But now it's going to be put into the Meta ad machine. So what can you tell us about this?
35:10 Speaker 01 Yeah, it was funny. I was talking with someone at Meta about this earlier this week. And they said, you know, I'm sure a lot of people already thought we were doing this, but we're just announcing that we're going to start doing this. And I thought that was a really telling comment about how people view privacy on Meta's platforms. But yeah, I mean, Meta has basically announced that, yeah, we're going to take everything you say to an AI chatbot, and we are going to basically have it. Affect the kind of ads that you see on Instagram and Facebook. Of course, there are some exceptions if you are talking about really sensitive issues. Meta has kind of called out a couple of ones, like if you talk about religion or politics
35:54 Speaker 01 or kind of sensitive health issues, then it's not going to suggest you ads based on that information. But this is, I think, a really kind of fascinating moment where the AI companies are coming out and saying this. And it is really, I think, revealing of their playbook in the coming years. You know, maybe these AI chatbots were once free products on their platforms, you know, that we've gotten to use for the last couple of years. But I think we're increasingly seeing that there are strings attached to using these free products. They are going to take your data and serve you ads on them. They're going to train more models based on everything you tell it.
36:35 Speaker 01 I think it might make a lot of people think twice about how much info they share with these AI chatbots. And I'm curious to see if it really affects, you know, how people use them as therapists and friends and whatnot.
36:49 Speaker 00 Right, it's good to see that there are some exemptions, but I can just imagine that the next scandal is going to be like, I've used the Meta AI chat, someone saying, I've used Meta AI chat as my therapist and now I'm getting ads in my feed for Prozac. Talk to your doctor about Prozac or something like that. Like, that would be bad.
37:08 Speaker 01 It does seem bad. Yeah, I think that there is also, oh yeah, I was just gonna say, I think that there's also this kind of, really, it's incredibly valuable data stream for Meta. And I think that, you know, this has probably been the plan all along in some sense is to, you know, take this profile that they've built via the AI chatbot and monetize it in some way. And we've been talking for so long about how, you know, you could put ads in a chatbot. But I think that people maybe have spoken less about just how valuable all that information is to serve you ads elsewhere. And Meta is probably the best positioned company to do that.
37:51 Speaker 00 The economics of this AI boom, do they make sense if they just boil down to serving us better ads, or does it has to be something more?
37:58 Speaker 01 I think that ads have been the biggest business on the internet for a long time now. And I have no, I don't think that's going to change. Like, I like, I think that that is going to continue being the case. I think it could probably be a bigger business than it is today. I think you could probably get more specific kinds of ads and AI. I mean, I share so, I and other people, everyone shares so much more information with an AI chatbot than you do with your Instagram feed with Google. And you have to think that, you know, these companies that already have this great profile on you, if they could fine tune it even more to show you just the exact product that you want
38:40 Speaker 01 in the moment that you want it, even better than they can today. I mean, there's clearly a lot of value there. So yeah, I do think that even if it just boils down to ads, I mean, that's, that's gotten Google and Meta pretty farce, so I wouldn't, you know, doubt that.
38:58 Speaker 00 Max, I'm, I'm getting depressed.
39:03 Speaker 00 So my take on this would be ads will be part of it, but it has to be substantially more than ads. They have to end up automating a good chunk of white collar work to make the valuations work. But certainly the companies are going to be much more sustainable if they can use it to be much more effective in advertising. So I think you're probably right there. I'm curious what the vibe is inside of Meta right now, given, well, the state of vibes and the new talent and the pivot or the move towards commercializing its AI technology. This is from the information. A Meta change on publishing research causes a stir in its AI group.
39:45 Speaker 00 In recent weeks, Meta Platforms made a change to its fundamental artificial intelligence research lab that rubbed members of the group the wrong way. FAIR, as it's known, would have to undergo additional review of its research before publishing it. The new policy was a tough pill to swallow for some members of FAIR, who perceived the change as a new restriction on the freedom that they had, that they previously had to share their research outside of Meta. Meta has spent much of this year remaking its broader AI efforts and it has pushed FAIR to contribute to Meta's products and spend less time on sharing research externally the changes. So bothered Yan Le Kuhn, an AI pioneer who co-founded FAIR in 2013, that he mused
40:28 Speaker 00 to colleagues in September that perhaps he should quit as FAIR's chief scientist. Max, what's your read on what's happening here? And do you think that we're just at the beginning of massive tumult within Meta's AI division because of the focus on products and because of this talent situation that they have?
40:45 Speaker 01 Yeah, I mean, this is a great report on the state of Meta right now. I think it lines up with things I've heard from people inside of Meta in previous weeks that basically the level of bureaucracy inside of Meta, specifically in its AI division, is really a problem for a lot of people. I think that this is something that new employees that have come in from OpenAI and Scale and whatnot have complained about, but it also seems to be leaking into FAIR. What's interesting is that FAIR has been for a long time now allowed to kind of operate
41:25 Speaker 01 independently very, you know, they could kind of work on these lofty, long-term goals, you know, sometimes wouldn't, you know, produce anything for five to 10 years, but they were allowed to kind of operate independently and they've produced a lot of great work over the years. I think that is increasingly in jeopardy as AI just becomes too important to Meta, where, you know, as this report states, I mean, they like, we need to be really careful about what we put out. We don't want to give our competitors an edge. And yeah, I think that there is a lot of moving pieces inside of MSL right now and
42:08 Speaker 01 people are coming, people are going, almost every week, it seems. I don't think that this will be the last story we see about, you know, the kind of tumult inside of MSL.
42:21 Speaker 00 Yeah, MSL Meta Superintelligence Labs. And it is interesting because I've written a lot about the formation of FAIR back in the early 2010s. And Zuckerberg was absolutely right in seeing how important AI was going to be. And it is interesting that they didn't have a high opinion of LLMs within FAIR for a while. And he probably feels behind the eight ball and trailing OpenAI, or he definitely feels that way, which is why he's spending this much money on talent. But question is, now that that bet has sort of proved really smart, how much you let a lab that's focused on future stuff continue to think about that as opposed to going full
43:06 Speaker 00 speed ahead on the present? So, speaking of the present, we have a new model from, and we're completely under the radar this week. Anthropic launched Sonnet 4.5, its best model for AI coding, you write in TechCrunch. It's capable of building production-ready applications rather than prototypes, a leap in reliability from previous AI models. And also, I think it can go autonomously for like 30 hours. So, Max, you wrote about this. What's the significance of Sonnet 4.5?
43:34 Speaker 01 I think the story of Sonnet 4.5 is really just Anthropic needs to get back in the lead in the coding space. I think that for a long time, as you've talked about on the show before, I mean Anthropic's models are really a favorite among developers, and they have been for the last year or so. And it's because they're very good at coding, they're very good at kind of agentially working on their own. So in apps like Cursor, in these vibe coding apps, a lot of people use them, and it represents a huge part of Anthropic's business right now. GPT-5, the model OpenAI released a month ago at this point, I believe, two months ago,
44:17 Speaker 01 it really changed that. People have talked to a Cursor, say that it completely leveled the kind of usage between OpenAI and Anthropic's models in a way that they hadn't seen for a long time. And I think that Anthropic, you know, for OpenAI, I mean, coding models are just a small piece of their business. ChatGPT is kind of the the big tuna for them. But for Anthropic, I mean, this is existential. Like, they need to lead in coding, and this is their attempt to do that. And whether Claude Sonnet 4.5 does that, it's on benchmarks, very impressive. But I mean, I've heard mixed things from people saying GPT-5 is really strong. It's still my go-to. So I'm curious to see if this puts Anthropic back in the lead.
45:04 Speaker 00 I was speaking with the CEO this week, who unprompted, talking about like backwards compatibility with AI and said, hey, all of my engineers, you know, went from Anthropic's models, and we ran towards GPT-5 when that came out, and then everybody ran back to Sonnet 4.5. So it is one of those things where you now you see this competition. And if I was an engineer working with AI code, I'd be thrilled because you have these two heavyweight, multi-billion dollar companies just trying to make my life easier at every possible turn.
45:35 Speaker 01 Yeah, I mean, it is an insanely competitive space right now. I mean, you you basically have to ship a new model every couple of months to stay ahead of the competition. And I think that this is a really tough business for Anthropic to be in, where it just has to execute at a very high level for, you know, the foreseeable future. Like, it can't really take a break here. It can't miss a step. I am curious to kind of see if this really changes the way people use AI coding models. I mean, Anthropic says that it's capable of building these production-ready applications that, you know, you can vibe code up an app, and it'll set up, you know, a database, and it'll become SOC 2 compliant, and, you know,
46:19 Speaker 01 all of these things. And I'm curious to see if this actually enables everyday people to start using AI coding models more, or if it just makes a developer's life easier. I lean towards the latter.
46:34 Speaker 00 I think the latter probably is it. Listeners and viewers, we have Mike Krieger, the head of product at Anthropic. He'll be on the show next Wednesday for our Wednesday show. So stay tuned for that. We'll talk to him all about Sonnet 4.5. All right. Final story to this week. Let's talk about the Apple forthcoming smart glasses. This is from Bloomberg. Apple's smart glasses are so important to it that it's shelving the Vision Pro headset, this smaller one. Oh, actually, actually, no, I'm gonna let me read the story. This might be bigger than I expected. Apple Inc. has hit pause on a planned overhaul to its Vision Pro headset to redirect resources towards a more urgent effort, developing smart glasses that can rival products from meta platforms. Smart
47:18 Speaker 00 glasses have emerged as a critical arena for tech companies which are racing to develop AI centric devices. Future designs could eventually challenge smartphones as a must have technology and Apple wants to be ready. The company is working on at least two types of smart glasses. The first one dubbed N50 will pair with an iPhone and lack its own display. Apple aims to unveil this model as soon as next year ahead of a release in 2027. Apple is also working on a version with the display, something that could challenge the just released Meta Ray-Ban display. The Apple version had been planned for 2028, but the company is now looking to accelerate development. This is this is a massive story. Thanks to Mark Germain from Bloomberg to for reporting
48:02 Speaker 00 it out. But it's fascinating to me that Apple, looking over at Meta, now sees glasses as so important that it's shelving a big redevelopment of the Vision Pro reportedly to go after. What's the significance, Max?
48:16 Speaker 01 Yeah, I mean, I think that honestly, Meta's Ray-Ban display glasses were so impressive to a lot of people. I mean, if you try to book an appointment in the Bay Area to go demo these, it's basically booked out for the next two or three months. I mean, you basically just can't even. Yeah, you can't even get an appointment right now. So and you have to demo them to buy them, which is a weird detail about them. But but the demand for them, I think, has been incredible. I think people were really fascinated by this new computing form factor. I think it's a lot more compelling than putting a pair of big goggles like an Apple Vision Pro
48:58 Speaker 01 or a Meta Quest on your face. I don't know about you. I can't stand the headsets, to be completely frank. I think they get hot. I think they're kind of uncomfortable. But the glasses are very comfortable to slide on. And I, for one, am really excited about this development. I think this is like a much better direction for Apple to go in than making a better Vision Pro. But of course, Apple is in a much better position to win here. I mean, they can connect it to iMessage. They can connect it to all the apps on your phone. So I'm not sure what claim Meta really has besides being first to market here.
49:35 Speaker 00 Right. It is interesting because you have these two things going on at the same time. Meta being first, like it's kind of, I'm sure Mark Zuckerberg is loving the fact that Tim Cook is running after him in product direction. Like, he calls the company Meta, then Apple develops Vision Pro. He builds glasses, then Apple builds glasses, you know? It's like, he must love that. But you're right. I mean, Apple has this tendency to be last and best, and or not, maybe not last, but it releases when it's ready and it's often best. The one thing I would say with Apple, it's, these are effectively AI devices, right? These wearables become interesting because of the ability to integrate AI into everything. And at the risk of saying the same thing I've said on the show for like five years already in its entire existence,
50:21 Speaker 00 if Apple's gonna play here, it's gotta get better at AI.
50:25 Speaker 01 It definitely does. I mean, I think that this product, more than any iPhone that Apple releases, will be an AI-first product that probably requires you to talk to an AI model to use it at all. So it does need to get a lot better. I have been hoping that Apple would just, you know, partner more seriously with OpenAI and Anthropic and, you know, just opening up their platform so that other AI models can really just, you know, power some of these features that Apple clearly just can't develop. But Apple has a lot of pride, so I'm not sure if they'll go down that route.
51:08 Speaker 00 I just want to give a quick shout out for Apple. Apple, I know I've been tough on you on the show. No apologies, but I will give you credit where credit is due. I recently picked up a pair of the new AirPods 3 Pro model, and I've never used better headphones in my entire life. They are crazy. The sound is amazing. Even though they're earbuds, the thing can really noise cancel. And it's now my favorite new tech thing I've had in a while.
51:36 Speaker 01 Wow. Have you tried the, there's like the AI translation feature in them? Have you tried that?
51:43 Speaker 00 Not yet. I need to figure out how to set that up. I think you need like to get the translate app and probably I'm on the 15 Pro for my phone. But my wife's family, or my wife and her family, they're all European. So I think that it'll be really nice for me on our next visit out there to be able to turn, turn this mode on and be able to communicate with their grandparents. So yeah, I'm looking forward to it.
52:10 Speaker 01 This is the brilliant part about Apple is that when it does release an AI feature, it immediately has millions of people who are so excited to use it and interact, you know, with their daily life. This is the thing I can't get over. And I'm sure Meta is acutely aware of that just, you know, the smart glasses may take a couple of years to come out. But when they do, I mean, that thing is going to sync up right to your iPhone. It's going to have, you know, all of your contacts in it. And I don't know, like, I just think that Meta can, has to really innovate in the next couple of years and race ahead to have a lasting chance here.
52:44 Speaker 00 Yeah, but the thing is, I guess it's now it's in the game. I, so I've had a chance to sit with Zuckerberg a couple of times. And the thing that I saw him the most frustrated about in our conversations was, was the fact that he has to go through Apple and Android in order to deliver his services. He hates just being an app on the phone. So I think he'll take the fact that he's in the game. But it's not going to be easy to run away with. So. All right, I want to end here. On your desk, if I'm not mistaken, right now you have the Friend pendant. Folks, I don't know if you've seen it, but friend.com is this AI wearable device that a company, they listen to your conversations, they give you updates and on your day.
53:29 Speaker 00 They're running a massive ad campaign in New York right now. Everywhere you look, there's these friend signs. And as Ari Papparo said last week on Twitter, that this is chaos marketing, science meant to be defaced because it's just like these white signs and, you know, unclear slogans. Long story short, you have one and I'd love to hear your experience with it, Max.
53:51 Speaker 01 Yeah, I do have one of the Friend devices. I got it a few days ago and I think that it is a really fascinating experience to try out an AI wearable. I will honestly say that, you know, like, there's a lot of things that Friend is doing, I think, really well right now. Like, I think that their marketing is honestly like, like one of the smartest marketing campaigns I've seen in a long time. Even if it is rage bait, I mean, it's getting so many people talking about this. It's incredible. And then the actual product itself, I think the packaging, I think the design, the experience of getting one of these things is very thoughtful.
54:33 Speaker 01 It does feel kind of like an early Apple product in the way it's packaged at least. And I know that the CEO, Avi, definitely puts a lot of attention into that. The experience of actually using it is a little bit unsettling in some ways. I mean, the device is, you know, it kind of has a snarky personality, kind of like you're talking to a 22 year old. It kind of has, you know, this, I mean, it has this always listening feature where it is, you can't turn it off really easily. Like, I mean, I've had some sensitive conversations with sources that I've just literally had to disconnect the device from my phone completely and like toss it in another room.
55:11 Speaker 01 Because and you, you also have to sign this like crazy privacy agreement with Friend where you basically have to agree that, you know, they can use your data for whatever, you know, that you can't sue them. You can't do a class action lawsuit against them. I mean, that, I think there are a lot of, a lot of strings attached to using a Friend device. It does make me wonder the whole experience about. How can there be a good version of an AI wearable like this? Like, this is a pretty strong starting point, I think. But, you know, what is OpenAI going to come out with? Like, you know, is everyone going to feel the same way about OpenAI's device as this? Or is there some way to make people feel a little more comfortable about it?
55:55 Speaker 00 And the output that it gives you, useful?
55:58 Speaker 01 I think the out, yeah, it gives you summaries of your day. It'll sometimes just text you like it's designed to be like a friend. Like, so it'll just, you know, like you could, you could be like, I was talking to someone and it texted me something about like something he said like, like, man, like, I can't believe Finn said that. Like, like, what, what was this guy talking about? And it does feel like it is trying to just get you to engage with it more. Like, it is, I'm almost feel like I'm being rage baited by this like AI chatbot that's always listening to me. And it is interesting. Avi has said that early users of Friends send like an average of 200 messages a day to it, which is a lot.
56:43 Speaker 01 And, yeah, I mean, I have like a lot of thoughts about using this device. I'm not sure if this is something that I will, you know, I'm going to be a power user of, or I'm going to use forever. But I think that it's fascinating to get, you know, a first experience with an AI wearable like this.
57:00 Speaker 00 Very cool. You're the first person that I've spoken with who's used it and can give us some insight into what it's like. And I have to say, I'm not running towards it right now, but I'm, I will eventually, I'm sure I'll eventually use one of these things, but, but I'm not running towards it. All right, Max, before you head out, why don't you tell folks where they can find your work and where they can find you online?
57:19 Speaker 01 Sure. So you can find my work at techcrunch.com. My name is Maxwell Zeff and I'm on X at at Zeff Max. And you can also find me on Threads and Blue Sky under the same handle.
57:34 Speaker 00 All right. Well, Max, as I mentioned, a big reader of your work, and this is the first time we've gotten you on the show. And I hope we are able to have these conversations many more times. So thanks again for all the great work.
57:46 Speaker 01 Thanks for having me. Would love to come back.
57:49 Speaker 00 All right. All right, everybody. On Wednesday, Mike Krieger, Chief Product Officer at Anthropic, is going to be here to talk about Sonnet 4.5. So we hope to see you then. If you haven't checked it out, I have an interview on the feed with Scott Guthrie, the head of AI and Cloud at Microsoft, who did have some things to say about the AI build out and why Microsoft seems to now be taking what he might call a more careful approach than some of the others out there. So definitely encourage you to check that out. All right. That'll do it for us here today. Have a great weekend and we'll see you next time on Big Technology Podcast.
58:21 Speaker 02 Now, more than ever, we need someone on our side. But Rob Whitman? Whitman's votes drove up costs, cut Medicaid, and risked 350,000 Virginians' healthcare to pay for tax breaks for himself and his donors. He's not on our side. Shannon Taylor will fight corruption, cut taxes, and bring down costs. This November, vote for Shannon Taylor. She's for us. Paid for by HMP, the House Majority PAC dot com. Not authorized by any candidate or candidate's
Previous Microsoft's Head of Cloud & AI on the AI Buildout's Risks and ROI — With Scott Guthrie Next OpenAI and AMD's Megadeal, Sam Altman and Jony Ive’s Bumpy Start. Meta vs. Apple Page updated October 12, 2026 · transcript and summary are machine-generated