← Mixed Signals from Semafor Media
Mixed Signals from Semafor Media
Instagram CEO Adam Mosseri on algorithms, AI slop, and how to beat TikTok
00:44:39
Loading summary
Transcript97 lines
- [00:11]
A
Welcome to another episode of the Mixed Signals podcast from us here at Semaphore. I'm Max Tawny. I'm the media editor at Semaphore, and with me, as always, is our editor in chief, Ben Smith. Ben, how's it going? Do you like our new theme music? What do you think about it?
- [00:25]
B
You know, I think it's great. I think it captures the sophisticated and yet engaged sensibility of this show that we. We told the AI to write for us.
- [00:34]
A
I have to stop you right there, because I don't want the audience to even think for a second that this was done by AI A human made those sounds that you were just hearing. The audience also probably picked up the fact that I am much more invested in the theme music of the show than you were.
- [00:47]
B
That's why you're the executive producer, Max.
- [00:49]
A
That is true. That is why I'm making all the big calls on the show. But this week, we are so excited. I didn't make this call. You actually did, Ben. You booked our guest. We are extremely thrilled to have on the CEO of Instagram, Adam Masseri. They have just rolled out this interesting new feature in which you can basically watch Instagram reels on your television. Now, Ben, this plays a little bit into something that we have been talking about a lot on the show, it seems like this year and recently.
- [01:19]
B
Yeah, you know, I've been trying to get Adam on for a long time because he's been right at the middle of the Facebook meta Instagram revolution in media. He ran newsfeed for Facebook when I met him. He's, you know, very close to Mark Zuckerberg, but he also comes out of design. And I always. He may take this as an insult, but I always thought he was better at talking to east coast media types than many of his Silicon Valley peers and seemed to understand where the media industry, where the news industry are coming from. And now he has walked, like, directly into our kill zone of is everything becoming television?
- [01:55]
A
Yeah, I mean, we're obviously really interested to ask him about those new features, but also just broadly what Instagram is in 2025. I feel like I've been on my own kind of personal Instagram journ. It used to be just like this thing that I had on private for me and my friends, and I would post pictures of my personal life that I didn't necessarily want everybody to see or just wanted only my friends to see, and, you know, kind of would post jokes or whatever. And now it's like my own personal branding platform. And I think they've torqued it to be that way. So we really want to ask him, like where Instagram is now compared to where it was five years ago and also where it's going. It's going to change a lot with AI and with the ways in which short form video has just exploded. So it should be a really interesting conversation.
- [02:40]
B
Yeah, at some level the question really is just like, what is social media? And really, is social media still a thing? We can ask him those questions.
- [02:47]
A
Yes, absolutely. We're going to take a short break and we'll be right back with Adam Mosseri right after this.
- [02:56]
B
In an era where AI is rewriting the rules and consumer behavior is changing by the minute, how do you stay relevant? You turn to the experts. Think with Google brings you insights from top CMOs, practical guides on emerging tech and strategies that drive real growth. It's like having a market marketing guru in your pocket. Visit thinkwithgoogle.com to become a forward thinking marketer. Your next game changing idea is just a click away.
- [03:23]
A
Okay, Adam, we are so excited to have you here with us. I wanted to start out with kind of a basic question, but one that Ben and I were both interested in, which is what is Instagram in 2025? We feel like it's changed a lot over the past few years and it's probably going to change a lot more of the next few. How would you describe what Instagram is in 2025?
- [03:46]
C
So the mission has been for a long time to inspire creativity that connects people. But how people connect with their friends, how people create, how people entertain themselves has changed a lot. So the how changes constantly. And I know that can be a lot sometimes for people who are used to things. One way, I think the most important things to understand About Instagram in 2025 is that the primary way that people actually share with friends is through messages. It's not through stories or feed. And the primary way that people entertain themselves and consume public content is video. So it's a lot of DMs and it's a lot of reels right now. And then the next biggest surface is probably stories, particularly for younger people. So broadcast sharing, which is a term we use for sharing with lots of people at once, is still a thing, but that's for most people primarily as stories pattern, not a feed pattern.
- [04:39]
B
Would you call this thing a social network?
- [04:42]
C
Yeah, for sure. I think one of our strengths is the fact that you have your friends there. The way I think about it is most media or attention businesses have Content that is differentiated, which is why you go, and then more content, which might be why you stay. So, you know, maybe you subscribe to Time Warner when I was a kid on the east coast to get to be able to get access to hbo. Or maybe now you subscribe to HBO to get access to a specific show for Instagram. That hook for a lot of people is friend content and their friends. So it might be sharing their biggest moments or sharing lightweight moments and stories, or it might be them just sending them funny comedy clips in DMs. But that's like a real motivator for people to show up and then they can stay and spend a lot more time on public content.
- [05:30]
B
So we had actually like sort of an etiquette question on this front from a mixed signals listener, Kyle Chayka of the New Yorker, who wanted me to ask you, should people still post about their daily lives on there or just DM their friends in private?
- [05:43]
C
I think posting about your daily lives is totally inbound, but it's particularly a stories thing. It's not so much a feed thing. Posting twice the feed in the same day as a normal person is a little bit sort of outside of the lines of Instagram etiquette. They call it double insta. Ing or double Instagramming.
- [06:04]
A
So, Adam, it's funny you bring that up because one of our previous guests on the show, Jonah Weiner, who writes the excellent newsletter Blackbird Spy Plane, he had a post a few months back about how it was very transgressive these days to do the kind of 2012 style daily posting just on grid and that that was like kind of a cool thing that people have been doing again.
- [06:24]
C
Yes, that definitely happens. I think what is old is new is again is definitely a bit of a pattern. We've seen this with like disposable cameras. What's a bit of a broader theme than daily life to grid is the aesthetic. Aesthetic has just changed a lot. People think of Instagram, they think of hyper polished, photoshopped, high saturation landscapes, et cetera. But the content that is, I think, driving the most cultural relevance right now is kind of the opposite of that. It's like unprocessed, much more raw, much less, you know, photo dumps. I think people in a world where they're inundated with hyper produced content are gravitating more and more towards content that at least feels more authentic.
- [07:08]
B
I've come chasing you to come on this show for a while, but you finally had some news to make, which is Instagram Rolling out an app on Amazon Fire will have channels like news, entertainment. And, you know, a huge theme of this show is that everything is becoming television. And like, that really sounds to me like, wow, Instagram is now becoming television. What's going on?
- [07:31]
C
So I don't know that I would say that Instagram is becoming television, but we should definitely go deeper into that. I think that we're just seeing how important TV is as a format, and it's just a place where people spend time. Right. And historically, we've been fighting with some pretty fierce competition for people's time on mobile phones. But obviously there's an immense amount of time spent on tv, particularly here in the us, but also in Western Europe, also in India, where we're not even really showing up. And some of our competition is showing up with force. So for us, it's a bit more about making sure that we're not ignoring such an important media or medium. And honestly, I think we're late to the game, so we've got a lot of catching up to do.
- [08:15]
A
Obviously, plenty of people are watching YouTube on their TVs and we'll get to that. But when did you decide that this was something that both that you wanted to do and that also that Instagram could do on television in a way that would feel fun and interesting and, you know, good for the brand?
- [08:28]
C
I guess we've wanted to do it for a long time. I think it's about when we really decided to actually do it. TV is difficult to build for. There's a lot of fragmentation for all these different devices. It's a completely new format. Obviously, for us, the input device is really difficult. So how do you. And we've all fought typing to our tv, you know, your passcode when you can't dictate, you know, so all these kind of fun challenges that we're not used to. We wanted to start on one or two platforms to start, but then quickly expand and we'll see. This is. Look, the first version that we're announcing now is really just that it's a first version. We're going to learn a lot. I'm sure we'll get a bunch of things wrong, but we're going to iterate quickly. The other thing I would say is over the course of the year, it became clear that, though I wish we had done this years ago, the technology's changing in a way that is going to make this a lot easier.
- [09:18]
B
I think of you, Adam, and maybe I realize that you might take this sort of as an insult but as sort of a. I love that preamble.
- [09:25]
C
Before you even tell me. I just want to talk to you, tell you how.
- [09:27]
B
But it does feel to me like the things that you have spent a lot of your career building. Recommendation engines, newsfeed. The structures around Instagram and threads are technological, but there's something about this, like, television experience which is leaned back fundamentally, that feels more like, welcome to the media business, Adam. I mean, do you feel like you've crossed a threshold here?
- [09:48]
C
If there was a threshold, I feel like I've been stepping over it slowly for a long time. I don't know if there's like a Rubicon, so to speak, but I do feel like this is gonna be humbling. I think that we're gonna learn a lot that we don't know, and I think that's gonna be healthy. I'm constantly concerned about our blind spots. What are the things that we should be concerned about or thinking about that we are not?
- [10:14]
B
I mean, are you concerned that what it's gonna be is that high production, scripted dramas beat out UGC and next thing you know you're competing with Netflix to buy the remake of the Office?
- [10:27]
C
I hope not. I really don't want to get into that business. I think the more likely, the more likely tension would be. I don't know that this will be the case, but we have been very explicitly and intentionally focused on short form video on Instagram. My belief is that connecting people with their friends and short form video can be really symbiotic, you know, because you see short form videos you want to send to your friends. For me, it's, you know, comedy skits I send to my brother or, you know, things about architecture or art that I send to my sister, she's a designer. And then we have conversations. So those two things are, you know, they help each other. And then I think short form and long form video can be symbiotic. You know, you watch a couple short form videos, maybe you watch a long one, you go back to a couple shorts. I do not know that connecting people with your friends and long form video can be symbiotic. I'm a little bit worried they're too far apart. You watch a 20 minute video which doesn't even seem like long, but in the Instagram world is very long. That's a lot of reels that you wouldn't have seen and sent to a friend, or seen a story from a friend, or sent a message to somebody you know or love. So I think that if this one way this might fail. We'd like to do these pre mortems, like, how might this not work before we even find out, is that it might turn out that maybe, like you said, we'll need premium content to work. I don't know. I think YouTube is not big into premium. And YouTube on TVs is much bigger than YouTube TV for TVs as far as we can tell. But it might be that we need long form video. And then if we need long form video, what does that mean for us? Because that's literally been a market we've explicitly decided not to enter historically.
- [12:00]
A
So before we actually get into YouTube, because I can just, I can feel Ben chomping at the bit to. To get into it because we've done.
- [12:05]
C
So much this year, he perked up.
- [12:07]
A
I want to ask, like just a very basic question here because I think a lot of our listeners are probably thinking about this. How do you envision people using this? Like, okay, so I've hooked up Instagram now to my television. I have an amp. We've got Amazon Fire. This is. This is awesome. Now what happens? We're just. We're watching some reels in a row. Why wouldn't we just do it on our phone? What is the hab habit supposed to look like?
- [12:27]
C
I don't think we know yet. But I have two different ideas in mind. One is much more like how you use Instagram, but more passive. So, you know, we all have been in the place where we want to sort of hack our mood or watch some funny videos or whatever it might be, and we sit back and relax and it might just be nicer on a bigger screen, you know, when you're home alone. The one that I also think will be less frequent but maybe more delightful and exciting and valuable would be something that was more social. I mean, I don't know if you do this, but I definitely have been in a bunch of situations where, like, I really want to watch a bunch of things with a friend, you know, so with my brother, it's usually comedy because we both just love standup. And I think laughing is just, you know, underrated with, you know, other people in my life. It's other topics where we sort of wanted to go deep and rabbit hole together on something. I think this is often isn't in the first version, but I think it could be pretty exciting to do that either in a direct way where you say, here, this is the kind of content I want. And as content search gets better, we could deliver on that. And then eventually, in a more personalized way where we were like, hey, you know, you two are in the room. Do you want a feed that's just based on both of your interests? I don't know if you've ever used Blend, but it's a pretty funny way to get a little insight into your friends interests.
- [13:39]
A
Do you see a world in which people are creating content for Instagram on tv? This was something that came up when we were talking to Neil Mohan earlier this year where he was talking about how a lot of their users on YouTube were creating content specifically that was meant to be watched on, you know, your television screen in your living room. Is that kind of the goal? Do you want to get to that place?
- [14:00]
C
I think that's not the goal, but I think that would be a sign of success. I think that if the volume of usage gets so high that people would actually bother to create content specifically for Instagram on tv, that would be a sign that we had reached scale. But I think that's step three or four. I mean, we have to get something out the door, then we have to make it good, then we have to scale it across all the platforms, then we have to figure out how to grow it. So there's a lot of steps between now and then, but that would be, I think, a nice signal that we had reached enough scale that it mattered enough to somebody to even consider doing that.
- [14:34]
B
I guess I have almost sort of the opposite question, Adam, which is like, it feels like there was a period and, you know, I think Max and I are both like, you are veterans of the first pivot to video for Facebook Watch, in which people were making bespoke contents for each platform. And at BuzzFeed, we felt like we had real expertise on what worked better on Pinterest versus Facebook or whatever. Now I think a lot of creators are treating these as identical channels in the same vertical. Video goes on. I mean, this is certainly true for us and we were pleased to see, well, you know, you subscribed to Semaphore on Instagram for a video. I do think we posted the identical video to TikTok and to Shorts.
- [15:10]
C
True story.
- [15:10]
B
I mean, that sort of threatens to commoditize you in a way.
- [15:13]
C
Yeah. So there's signals in both directions. So we care a ton about creators. I mean, the whole identity of the brand is really around creativity. We have seen that creators every year seem to diversify more and more gradually but steadily. So they use more and more platforms. There's less and less creators who are just completely loyal to one platform. We have heard both. We have heard from some people that they have a lot of success in just reposting the same content across platforms. We've also found other creators who really believe deeply in the fact that they've got different audiences on different platforms and they find a lot of success in catering their content to those audiences. Sometimes it's just slicing the same content up, but often it's just completely different content. I think it probably varies by vertical and I don't have a great sense of which verticals that's true for and false for and why. But we def. I definitely hear both a lot. But it is very true that creators are becoming savage at the fact that it's valuable for them to not have all of their eggs in one basket, to not be beholden to one algorithm or one company or one risk of being deplatformed, and are trying to make sure that they have just a diverse business, which I think just makes sense in terms of the risk to us. Yeah, there's risk there. There's things that both TikTok and YouTube are better than us at. In the context of video creation and distribution, I think TikTok is the best in the industry at breaking new talent from small names. I think YouTube is the best in the industry at paying out creators at scale in a sustainable way. I think one of our unique strengths though is we are often not only a video creator's distribution platform, but also a consumption one. I think a lot of creators use us as a home base to consume content. Not just that's public, but also from their friends, to message people, to network. I can't tell you how many creators I've met from how many different industries that have made really important professional connections because they follow people whose taste that they love and then end up collaborating on something. From musicians to journalists to, you know, film to authors. So even non visual artist.
- [17:26]
B
Are you in bidding? Do you get in. Do you get into bidding wars for creators with either TikTok and YouTube or with Amazon and Netflix or with CBS and ABC? Do you feel like you're in a. A competition, whether monetary or otherwise, for their attention and time?
- [17:43]
C
I can't think of any bidding wars where there's literally something that we are bidding on. But we are certainly in a lot of competition with particularly YouTube and TikTok for Creators Time and attention. And I think of creators as particularly important because I feel like across a lot of industries, power is shifting from institutions to individuals and so they seem to be getting more and more important over time. They punch above their weight in terms of engagement as well. But just think about how relevant athletes are relative to the teams they play for compared to 10, 20 or 30 years ago. Relevant journalists are relative to the mastheads that they work under compared to 10, 20 or 30 years ago. Now, obviously AI is a big topic and there's a lot of concern around being disintermediated or competing with AI. But I think people fundamentally really want to hear from other people. And so we still feel like creators are one of the most important audiences for us across industries.
- [18:40]
A
Well, we have a lot that we want to get to with Adam Mosseri, but we have to take a short break. We'll be back right after this.
- [18:53]
B
In this week's branded segment from Think with Google, I spoke to Google's VP of marketing, Josh Spanier, about whether there's space for CMOs in the boardroom. Usually when you look at a corporate board, you see a lot of former CEOs, CFOs, maybe a legal expert. But Josh, you at Google have been doing some research which suggests that there is a blind spot when it comes to the composition of corporate boards.
- [19:15]
D
I've been seeing this fascinating research looking at the number of boards which have a marketer, someone with deep marketing expertise, or a former CMO on the board. Those companies that do have a 6% higher growth rate over the study period than companies that don't. Here's the challenge. Here's the opportunity for marketing. Just 16% of boards in the United States have a marketer or deep marketer on their board. 84% of companies don't. That's a real gap. That's the opportunity that marketing can help with.
- [19:47]
B
I suspect cmos will be glad to hear this data, but what specifically is it that makes them valuable?
- [19:52]
D
CMOs have always been the voice of the customer. Inside corporations. They are closest to the customer. They understand them to a deep level. And in this moment of unbelievable transformation in how those customers are discovering your products, how they're consuming their media, how they're changing their behavior. Having a CMO who has got that insight directly at the highest level of the company is incredibly valuable. But second, marketing is sitting at the intersection of AI and nearly every department with the company. How AI is transforming your actual products, how AI is transforming your logistics, how AI is transforming media and advertising and every part of what's going on. Having a CMO who actually understands those technologies and can speak to the whole breadth of the organization at the board level is really valuable. And it's changing growth trajectories.
- [20:41]
B
Is there something cmos ought to be doing to make themselves more attractive board candidates?
- [20:45]
D
They have to speak the language of finance. They have to do all the things that CMOs are meant to be doing. A good place to start is over@thinkwithgoogle.com where they can get the latest insights and learnings on AI transformation and bring those insights collected across Google and across hundreds of companies companies inside their organizations.
- [21:02]
B
Thanks, Josh.
- [21:03]
D
Thanks, Ben.
- [21:13]
A
You were on Peter Kafka's podcast a few months ago and you said something which was interesting to me. You said that TikTok was more aggressive on data in general. And just now you mentioned that TikTok is good at kind of breaking new stars or new people. What did you mean by more aggressive on data? And why do you think TikTok tok has been better at kind of breaking new people? What is it specifically, do you think?
- [21:36]
C
So those are related, but two different things happening on data. I mean, I'll give you some examples of things that they do that I don't think anybody else in the industry does. If you message a TikTok video to a friend that creates a unique link so that they can track it and you can message in any other app, you don't have to use TikTok. So if I messaged Ben an encrypted message with a TikTok link in WhatsApp and he opened it and it opened up his TikTok. TikTok knows that I sent it to him and it recreates a message thread in TikTok but as if I had messaged Ben directly on TikTok.
- [22:10]
A
Wow. Wow.
- [22:11]
C
That is a very aggressive thing to do in an encrypted messaging world. Yeah, they are all in on those types of tactics. You know, that's about data, but that's also about growth. On breaking talent, I think that they deserve a lot of Credit. Even before ByteDance bought Musical Ly, Alex and Musical Ly were very leaned in on this idea of not reinforcing the existing establishment, but helping new talent grow. In the world of ranking, there's something called exploration based ranking which is just more difficult. It is much easier for me to find something that I think you're going to like that I know a million other people like than to find something that I think you're going to like that you know, only like 100 people have even seen. But if you do the latter, more interesting things happen. And I think you really are kind of living up to the whole promise of the Internet and niche interests. So they did a lot. And I don't need to go into the details unless you want me to on exploration based ranking systems really early that the rest of the industry has caught up on.
- [23:14]
A
Okay, this is what I wanted to ask about because I'm very knowledgeable about exploration based ranking in the sense that I learned about it an hour ago when I was prepping for this interview and listening to you on another podcast. But Ben, do you know what exploration based ranking is?
- [23:27]
B
I do not.
- [23:27]
C
The basic idea is that you want to give every piece of content a chance, regardless of whether or not it's from a big established account or not. It is much easier for us to drive up engagement by showing a lot of photos and videos from Selena Gomez and the Rock because they're so big and we know so many people like them. It is much more difficult to figure out that you happen to be in like into Afro Beats and you happen to be into like, you know, crochet scarves, you know. So the way it works is essentially almost like a little competition where you reserve a certain number of impressions to try things out. And you do this in different ways. But I'll start with the content way, which is you try to get every piece of content or some subset of everything. If you can't get everything, you know, 100 views, call it, and you see how they do, even if you had no reason to think it would be interesting to those individuals. And then you take a look at all of the things that you auditioned and you see what did much better than you expected. And then you graduate those to 1000 views with another percentage of your impressions that you reserve for these auditions. And then for the things that do disproportionately well in that 1,000, you graduate again to call it 10,000 views and so on and so forth, which is how you can see something from someone with 50 followers, get millions of views in a couple hours because they just created something that happened to wildly outperform everything else that day. And what is great about this, you allow new entrants to become to have hits. Not just the people, not just the sort of established largest names, but you know, it comes with a lot of volatility. It means reach is much more predictable. It completely relies on unconnected distribution. So seeing things from accounts you do not follow, which eats into the value of follow graphs, which is can be concerning and we get a lot of complaints from both the creator side and the consumer side, but also can be really beneficial from a Creator perspective where you don't want a small number of creators just to overly dominate all the impressions. You want to give new people, new ideas, a chance and then from a platform perspective you have a chance to earn some real sentiment from these small creators who make it on your platform. It's much more, I think more practical to try to break the next generation of creators and to pull an existing generation of creators over from another platform. So that's part of, that's one type of exploration based ranking which is essentially this audition system that TikTok really pushed the furthest, the earliest but now is very standard on Instagram, on YouTube, on Facebook and others.
- [26:04]
B
You're being so kind of nice and even handed with your rivals here. Like what do you think is TikTok's Achilles heel? How are you going to beat them?
- [26:10]
A
Them?
- [26:11]
C
Ooh, it's going to be hard to beat them.
- [26:13]
A
I think that the United States government.
- [26:16]
C
No, yeah, I mean that, that's a whole other conversation. I mean look, I think their rate of progress has slowed down and I think in part of that was the distraction with all of the regulatory and policy concerns. So I don't think they have been making progress as quickly as they have in previous years, which gives us some a chance to catch up on things like core ranking. I think we monetize way better than them, particularly if you look at not just revenue but profits. But their revenue globally is doing really well. But on the profit side I think we're pretty far ahead of them. Here's a big question is like they are very much applying lessons they've learned in China to the rest of the world and there's good and bad are pros and cons are to that. So one is the pro is they had a massive faith in the short form video format long before anybody else did musically. And the early versions of TikTok were not retentive and they spent billions of dollars a year dumping into promoting it. And it seemed crazy to the rest of the industry. But their faith from talking to people at their competitors in China came from the fact that they had proven the market in China and they were confident it would exist elsewhere and they just needed to keep pushing to get there. So on one hand that's really interesting, right? That's them seeing into the future in a sort of another way. The flip side is, you know, the super apps which are very popular in China are not popular in the same way outside of China and I think they're turning into a Chinese super app. TikTok is. And that may or may not work outside of China. And so, you know, now are they overly projecting one win on the next or am I just, you know, a naive, out of touch Westerner who's underestimating where the future is going to be? Like, I don't know. But that's, that'll be interesting to see how that plays out. But that I feel like Instagram's too complicated. And I love TikTok. It's a great app, but it is really complicated at this point and increasingly kind of crazy to get back to.
- [28:12]
B
This question about everything becoming tv. Like that thesis, which in Derek Thompson had a really interesting essay about it that we talked to him about recently is partly about kind of like literally staring at the big screen, but it is also about a shift into kind of like a visual and oral culture that I think TikTok led the way to that kind of more passive consumption. And I think in Derek's sort of concerning view turns us back into like pre literate cavemen. And I wonder like, does this concern you, Adam?
- [28:39]
C
I mean, it's hard to say no when it's framed that way. I mean, I like to think that. And we've tried on Instagram to make video more lean in than a lot of the competition, right. So we try to bring your friends there, right? So like you can see who else has liked something. You can sends are a huge part of Reel's success in a way that they weren't part of Facebook's dive into video in 2014 and 15. So it's actually grown the amount of people actually see each other and talk to each other, even though it's generally a laid back experience. We've got the dedicated Friends feed in the Reels tab, we've got Blend. We've been trying to create experiences that are a bit more social, a little bit more participatory and a little bit less lean back than like just scrolling for an hour or watching a Netflix show for an hour. But you know, we'll see. Maybe that's a smaller market.
- [29:31]
B
I thought you were gonna tell me that Threads was gonna save us.
- [29:34]
C
You've known me for long enough to know that I've tried really hard never to over promise and under deliver. So that's not the kind of thing that would come out of my mouth. Threads is fun cause Threads is about ideas and perspectives in a way that you know, Instagram is about creativity and videos and photos. So Threads does seem to have detached.
- [29:52]
B
From the hyper hyper polarized text based platforms. You know, X, which is hyper polarized in a lot of now, Blue sky, which is very polarized also. How do you think about that landscape?
- [30:04]
C
Okay, a few different things. One thing that really helped us is out the gate saying we wanted to be not like Kumbaya, but just a more like civil platform. You know, a platform for all perspectives implies that people can share things without getting destroyed for sharing them. And we try to build tools to back that up, not just in ranking, but, you know, you can turn off who can reply, you can turn off quote posts, you know, giving people hidden words, giving people some meaningful, small, powerful tools. Because we always knew Threads was going to be like Twitter in the sense that a small number of people were going to dominate the consumption. Right. It's going to be for most of us to enjoy a select few debating or sharing their perspectives in a somewhat performative manner. And so we thought building tools for that select few would matter a lot because you don't need that many people to use them for. To be valuable. You just need the people who actually post a lot to do it. So going out the gate trying to say we wanted to be civil, I think helped set the tone. I think X has done a couple things that have helped. Obviously they've leaned a lot into things that have been much more divisive, but also they've gone a lot after video over the last year or two. I think they're a little enamored with time spent and I think they underestimate the costs of video, so that provides more space for us. And then Blue sky is interesting because they're. They're the most zealot on the sort of open architecture side. Not that it's all academic. It's not. But they are heavily invested on their sort of academic side, on the theoretical. And then they maybe just based on our positioning and X's positioning, have ended up with a relatively small but very frustrated user base. So it's, you know, unless you have one particular worldview, it's pretty tough place to be.
- [31:56]
B
Yes, it seems like Threads, though, is doing fine, but at the cost of a certain kind of relevance. There aren't. The things that are being discussed in cable news are rarely posts to Threads, they're often posts to X, they're often posts to Blue Sky. Do you miss that? I mean, you annoyed journalists early on by saying you didn't really want us over there.
- [32:16]
C
Yeah, I know. I really made so many friends. Well, we gave up on that in a spectacularly public fashion earlier this year. There's a ton of news on threads at this point, floodgates are open. I think we don't yet have the cultural relevance we aspire to, but I don't think it's because of news necessarily. I think some of it's just old habits die hard. So if you look at different verticals, you can see there are certain areas where we are either leading, there's certain areas where we're behind, and certain industries where we're basically nowhere compared to some of the competition. And so the question is, how do we close that gap? But you know, like for instance, like NBA Twitter, a really long standing, deep community, we're actually making really good headway. NBA Threads had its own set of founders at the beginning who weren't big names and they have a ton of investment in that community. But now that you're seeing the players start to come over too, there's this interesting interaction between the players and the threads specific community that's growing, but still we have certainly not toppled NBA Twitter. That is a deep, long standing community. So it really varies community by community. But we think about this a lot. How do you measure cultural relevance? How do you define cultural relevance? How do you.
- [33:27]
B
And do you want it? Because I almost think sometimes these days cultural relevance is more trouble than it's worth.
- [33:33]
C
Like, it means he's just baiting me. You want it, we want it. But just to be super clear, there's a lot of cultural relevance that isn't the primary polarized political food fight, you know, where is the conversation about the playoffs for the NBA happening? Where is the conversation about the outfits on the red carpet at the Met Gal happening? Where is the conversation about the up and coming musicians, you know, like Cameron Winters happening? That kind of cultural relevance I don't think comes with the same cost as whatever the top headline, as combative as possible war culture, war of the day is, which I think is just gonna happen to us. And so we have to manage it responsibly but not pretend like it's not gonna happen to us.
- [34:23]
B
A lot of the most divisive stuff on X in particular seems to be AI generated. How are you thinking about AI generated content and slop on your platforms?
- [34:33]
A
Wait, Ben, can I ask a more pointed version of that question too?
- [34:36]
C
Ben's never gonna say no. Ask a less pointed version of my question.
- [34:39]
B
Just ask him more.
- [34:39]
A
But I wanna take it no, but Ben is leaned back here. I want to add, like, is AI content good for Instagram or not? Because, like previously, this was a place where you would interact with your friends. And it feels like it would change Instagram pretty radically for it to be filled with that kind of AI slop that Ben's talking about.
- [34:56]
C
So, in typical diplomatic fashion, I'll say that there are good and bad consequences of it, but I don't want to be vanilla. And the answer here is I just don't really think the primary type of AI content on the platforms is what people usually mean when they say AI content. They talk about AI slop. They're talking about fully synthetic content, usually really bizarre, weird, where like physics gets strange or people's arms bend in weird ways. But the truth is, I think there's going to be more content that is hybrid, that is AI is used to augment or enhance in some way than is purely synthetic. I think we're kind of already there because there's tons of ways in which AI, I think, can help creators make more content or make better content. Now, if you're asking about the purely synthetic content, I think it's going to be. I think there's in any kind of content, AI or not, there's going to be slop and non slop. Right. I think an important question, which is, are we doing a good job connecting you to the valuable or content that you would find valuable or interesting, whether it's about knowing that it's AI generated or not, or you're just interested in the topic or not, because it's a new type of content. And so naturally our systems aren't as versed in it, so we're not as good at ranking at it as we should be. So we're working hard on being able to identify AI generated content at scale and then just even just assess, like, how good are we at ranking it? This is tough because as the generative models get better and better, it's going to be harder and harder to detect that it was AI generated. So it might be a bit of an arms race, but it will be, but we'll see how we fare. But I just think there's a ton in the middle. Like, every time I post now, I have it translated into Hindi, Spanish and Portuguese. So if you go to one of my posts on Instagram, you can just switch it and see me speak Hindi. And it sounds like me and the lips look like I'm actually speaking Hindi. And it's by far our largest user base. And now obviously only about half the people in India speak Hindi, but that's a big number of people now they can actually understand me when a lot of them couldn't before, like, that's AI generated, but that's kind of exciting. Whereas, like, the slop that you have in your mind when you ask that question probably is something that you would wish you had never seen and you can't take out of your brain.
- [36:58]
A
I've seen some mesmerizing stuff we won't get into because we're almost done here, but because you have to go here. Adam, we did have a final question that we wanted to ask, which is, where do you think that Instagram will be in the next five or 10 years? What will it be?
- [37:12]
C
So in the next five, one of the big differences will be that a lot of the technology and some of these foundation models are going to allow us to allow you to much more proactively get your hands in there and shape it. Right? So, like, we launched your algorithm last week, which only applies to reels to start, but that allows you to see what we think you're interested in, to just. Just tell us when we're wrong, add things, remove things. But I feel like all the surfaces across Instagram with some of these new technologies should evolve in a way that you can kind of touch metal, where you can kind of go in there and just tell it what you want or make it what you want, whether that is shaping feed to give you what you're interested in, or going deep in search to understand something or creating something and being able to really mold it into what you want more proactively. I think that's gonna happen over the next two to four years. You'll start seeing it happen now, but I think it's gonna be. We feel fundamentally different in two to four years in a way that I hadn't foreseen only a couple years ago. The 10 years out is much crazier, because that's when I think the form factor, you know, goes to a place that looks more like glasses, probably becomes more important than phones. And then for a visual platform in a medium where you probably want most of your interactions with your smart glasses to be audio only, as a micro, much bigger, more open question that I think we need to think about.
- [38:37]
A
Well, we'll have to have you on again, then, to delve into more of the specifics.
- [38:41]
B
But Adam's gonna have a lot of time on his hands in 10 years, when the visual medium has gone away, I'm gonna be.
- [38:46]
C
Yeah, my kids will be grown, and I'll probably be cooking and reading a book and, like, reading. You know, reading, you guys, what you have to say on what's going on in the world.
- [38:54]
A
Well, we really appreciate that and this was a really great conversation. So thanks, Adam. Thanks for taking the time.
- [38:59]
B
Yeah, thank you for joining us, Adam.
- [39:01]
C
Appreciate your time. You guys are great. Thank you so much.
- [39:04]
B
Foreign. I want to tell you about Think with Google. Its marketers. Go to spot for insights from top CMOs, practical guides on emerging tech and strategies that drive real growth. You can learn how AI can help level up your marketing and address your biggest challenges across creative media and measurements. Whether you're a marketer looking for new ways to reach your audience or a curious leader wanting to get inspired, Think with Google is the place to be. Learn more by heading to thinkwithgoogle.com.
- [39:45]
A
So we are recording this the day after our interview with Adam Mosseri. We've had 24 hours to ruminate. Really let what he said sink in. Let's tackle the big question first, which is what do you think about his pushback on your thesis and the thesis of others who have come onto this show that everything is television? This is the thing you hear from all the smart media types these days. Everything is television. Derek was on here. You're a believer in this. Adam kind of didn't believe that. Said he didn't believe that while rolling out a promotion for this new Instagram Amazon Fire TV deal collaboration. What did you make? Do you believe him? Is the action different than words? What do you think?
- [40:28]
B
You know, it is so interesting because I think in some sense what's going on is that everybody is seeing this pull in the same direction. I mean, in this case just toward a large screen television format, and yet everybody is trying to resist it in a different way and keep some differentiation. Right. Each of these companies, and he sort of acknowledged this is running the risk that everything is becoming undifferentiated. And if everything is undifferentiated, the platforms lose a lot of power. If I'm just putting the same video in every place, which by the way, we in fact are. And so what struck me was that his impulse was to resist long form and to believe. And I think with a lot, with really good reason, that Instagram's strength, particularly its strength against TikTok is messaging, is personal connection, where TikTok is really a totally passive experience. And I thought that was so interesting that he said, well, personal connection and short form video go together. You send each other short form videos and then short form video and long form video go together kind of for obvious reasons, it's an evolution of the form. But that personal connection and long form video are too far apart and that long form video breaks Instagram. I don't know, I think that was just a very interesting observation. But it is true that even as like social media as we knew it has vanished. Meta's defense against its competitors has been the social graph.
- [41:47]
A
Yeah. And I think the other thing that Instagram benefits from is the fact that it's for a lot of people, their homepage almost like the homepage of their Internet lives. And that's why he, he mentioned that the thing, the leg up that they have on other platforms is that like they're big creat the celebrities or you know, the just people who are putting, you know, going viral all the time on Instagram are using it as well as consumers. Right.
- [42:10]
B
You know, it's funny, when I asked about threads, I almost thought he'd be like, ah, that was a fun experiment, nevermind threads. Because honestly from, in my own case I've kind of wandered off threads. What did you make of his continuing enthusiasm for threads?
- [42:23]
A
I found it to be interesting and consistent with what he said in other places. It's interesting. They've created like, like the second most popular kind of platform that's accomplishing what used to be Twitter, kind of behind Twitter. And I think it kind of seems like what they're doing is kind of like hanging around in the background until, you know, maybe Elon messes with Twitter again and there's this whole big kind of surge and they don't, they haven't really invested that many resources in it by meta standards. They said they have like what, like a hundred or something, something employees as a part of threads. So I kind of got the sense that if you're meta, you can really pour a lot of resources into a lot of these types of things. They realize that they want to, they want to capture the kind of like fervent tastemaker, you know, thought leader communities that exist still and are hanging out on the kind of decaying Twitter. They want to slowly kind of siphon those onto threads. And I think he rightly diagnosed the fact that Blue sky, while it has this kind of fervent fan base, that user base is also alienating to the vast majority of people who I think might wander back onto threads. Ben, are you going to go home and install Instagram on your television?
- [43:32]
B
I have a exotic projector situation on my TV rather than Amazon Fire, so I'm going to have to wait. How about you, Max?
- [43:39]
A
I just, I don't see the use case for it yet, but they've proved me wrong before, so I guess. I guess we'll both have to wait and see.
- [43:45]
B
Let's do it.
- [43:49]
A
Well, that is it for us this week. Thank you so much for listening to another episode of the Mixed Signals podcast from us here at Seminar before. Our show is produced by Manny Fadal and Josh Billenson, with special thanks to Chad Lewis, Rachel Oppenheim, Anna Pizzino, Garrett Wylie, Jules Zern and Tori Kaur. Our engineer is Rick Kwan and our theme music, our new theme music that Ben is so excited for is by Steve Bone. Our public editor is the algorithm, whether TikToks, which is better at using your data, or Instagrams, which is getting better at using your data. As Adam Szeri said, maybe we're not.
- [44:23]
B
Good enough at using your data yet. And for for that reason, please use our data. Please f follow us wherever you get your podcasts and subscribe to us on YouTube.
- [44:33]
A
And if you want more, you can always sign up for Semaphore's newsletter, which is out every Sunday night.