--:-- Erik Torenberg: I think in the future, you're going to see consumer spam to be like food, rent, software. All of the aspects of our lives are going to be intermediated by the models, and we're going to pay for that.
--:-- Justine Moore: The definition of what a companion is has evolved so quickly from either a friend or a girlfriend to anything, any sort of advice or wisdom or entertainment or counsel you could have gotten from a human before. Maybe you just need to feel connected to something. It doesn't need to be human.
--:-- Bryan Kim: I think it's been a puzzle to me what the first AI social network is going to look like. To work, a social network has to have real emotional stakes.
--:-- Erik Torenberg: We're living in this early era of AI where velocity is the most. For decades, consumer tech followed a similar beat. New platforms, new behaviors, new breakouts. Facebook, Twitter, Instagram, Snap, TikTok. But lately, that rhythm has changed, and so has the nature of what we mean by consumer.
--:-- In this conversation, we bring together A16Z's consumer and AI investing minds to ask, what is the state of consumer in the AI era? You'll hear how creative tools like Midjourney and VO are reshaping expression, how voice is becoming the new interface, and how companions—AI ones—are filling in social white spaces. We talk about retention and revenue curves, defensibility beyond network effects, and why velocity might be the new mode. We also get speculative. What happens when AI knows you better than your friends do? What does the next social platform look like? And are we heading towards a world where software, not shoes, not handbags, is the new luxury good?
--:-- This episode is about new form factors, new business models, and a new definition of connection. Let's get into it.
--:-- As a reminder, the content here is for informational purposes only. Should not be taken as legal, business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any A16Z fund. Please note that A16Z and its affiliates may also maintain investments in the companies discussed in this podcast. For more details, including a link to our investments, please see A16Z.com forward slash disclosures.
--:-- [Discussion begins]
--:-- Erik Torenberg: It seems like every few years there was a breakout starting from Facebook, Twitter, Instagram, Snap, WhatsApp, Tinder, TikTok. Every few years there was this new paradigm, this new breakout. And it feels like at some point a few years ago, it just stopped. Why did it stop, or did it stop? Would you reframe how we should think about that? And where do we go from here?
--:-- Justine Moore: I would argue probably ChatGPT was a huge consumer outcome and winner in the past few years. And we've also seen a bunch of other ones in various other modalities of AI, in image and video and audio companies like Midjourney and 11 Labs and Blackforce Labs, now things like Kling and VO. Weirdly, though, a lot of them don't have the same social or traditional consumer dynamics that you mentioned. I think because AI is still relatively early and so much of the new products and innovation has been driven by research teams who are so good at training models, but historically have not been amazing at creating the consumer product layer around them. So I think the optimistic view is that the models are now mature enough and many are available either open source or via API for people to build great, more traditional consumer products on top of them.
--:-- Bryan Kim: It's interesting that you asked that question because I was thinking about the past 15, 20 years, where, as you said, Google, Facebook, Uber, all the names. And it's interesting because when you think about Internet, mobile, cloud, everything together, there were all these amazing names. And you asked, has that slowed down? I think the cloud, mobile, all that had a lot of maturity baked in. The platform was around for like 10, 15 years. Every little nook and cranny has been explored to some extent. The changes that people had to adopt was Apple coming out with new features, as opposed to changes that people need to adopt now, which is the underlying relentless model updates. So I think that's one difference.
--:-- But the other thing is, again, Justin, you touched on this, but if I think about the past historical winners, there's the information area, like Google of the world. And now I think ChatGPT is certainly doing that. And there's a utility we missed out on, like Box and Dropbox of the world. There are more consumer, prosumery things that people use where we also see a lot of the companies attracting and going after that use case. Expression, creativity, same thing. The creative tools are endless and that's happening. What I think is missing potentially is connection. This social graph, this thing hasn't been rebuilt on AI yet. That may be just a white space or something that we just continue to see what develops there.
--:-- Anish Acharya: It's interesting because Facebook's almost 20 years ago at this point. The companies that you mentioned, Justine, aside from ChatGPT and OpenAI, are they going to be around 10, 20 years? What is the defensibility of the companies we're talking about? And also the use cases of all the companies I mentioned, are they going to be disrupted by these new players? Or in 10 years from now, will they continue to be the mainstream application for all those use cases that they serve?
--:-- Olivia Moore: I mean, you could argue that ChatGPT has got way higher business model quality than the analogous consumer companies from the last product cycles. Their top SKU is $200 a month. At the top, Google consumer SKU is $250 a month. So sure, there's a question of defensibility, networks, all these other things, but that might have been a response to the poor business model quality that would have occurred if you didn't have those things. Now you can just charge people a lot of money and perhaps we've been overthinking it previously.
--:-- Erik Torenberg: Yeah. There was poor business model quality, maybe stronger retention or product market or durability. You had to have a story for how this was compounding enterprise value in the absence of just making money right away. And now these models and these companies are just making money right away.
--:-- Anish Acharya: Yeah. I think the other thing is, Justine, you talked about this, all the foundation models are kind of pointy in different ways. So you could say, look, Claude and the ChatGPT horizontal model and the Gemini model, aren't they interchangeable? And doesn't that mean price pressure? But different people use them for different things and it seems like they're raising prices, not lowering them. So I think when you zoom in a little closer, you see that there are some interesting defensibility dynamics that are already there.
--:-- Bryan Kim: The increasing price, not decreasing, is an interesting point because monetization is clearly a different thing from previous era to AI era, especially for consumer companies. They're making money right away. One thing that's always on my mind, and Olivia, tell me if you think this is not correct, but the retention—when we talked about retention on the consumer subscription model before AI, I don't know if we actually tried to make a differentiation between unique user retention and revenue retention because they're kind of the same. You don't get to change pricing that often. You don't get to upgrade. It's the same thing. As opposed to now, we make a very clear differentiation between unique user retention and revenue retention because people actually upgrade. They have all these credits and points, they actually have overages that they end up spending. So you actually see revenue retention being meaningfully higher than unique user retention, which again, I haven't seen that before.
--:-- Olivia Moore: Yeah. I think before, the average consumer subscription was maybe $50 a year, if that. That was kind of a lot. The best in class consumer products would charge that. And now we have people very happily paying $200 a month. And even saying in some cases that they feel like they're being undercharged for that, or they would pay more.
--:-- Erik Torenberg: Can we explain that? What value are they getting such that they're paying more for it?
--:-- Justine Moore: I think it's doing work for them. Consumer subscriptions in the past were around things like personal finance, fitness, wellness, entertainment. But they were things that ostensibly would help you help yourself, entertain yourself. But you would have to invest a lot of time to get the value from them. And now with products like Deep Research, for example, that could replace 10 hours of generating a market report by yourself. That kind of thing is easily worth, I think, for many people, $200 a month, even on one or two generations.
--:-- Olivia Moore: I think things too, like VO3, people are paying $250 a month.
--:-- Justine Moore: Unfortunately, I have since—it's a limited, the $250 per plan.
--:-- Bryan Kim: You're the whale, Justine. You're the $1,000 a month.
--:-- Justine Moore: I've since charged several more of the $50 credit packs. So I probably spent, if I had to guess, over $500 in the first few weeks of VO3.
--:-- Bryan Kim: That's why revenue retention is higher.
--:-- Justine Moore: Exactly. And I'm happy to pay that because it's like you have this suddenly, it feels like a magical mystery box that you can open and get whatever video you want, only for eight seconds. But it's incredible. And the characters can talk and you can make amazing things that you can share with friends, make personal memes of someone delivering a message to your friend with their name in it. Create full stories that people are posting on Twitter and Reddit and all of these different places. It's like nothing we've seen before in terms of what consumer products can actually do for people.
--:-- Anish Acharya: It seems like every part of consumer discretionary spend is going to be overtaken by software. And I think in the future, you're going to see consumer spend to be like food, rent, software. And that's kind of where we're going, which is Dean speaking to.
--:-- Erik Torenberg: And can you give some examples of that?
--:-- Anish Acharya: Well, a lot of it is what Olivia said. I think all the entertainment is being subsumed by it. A lot of the creative expression work that you would do outside of software is now being subsumed by it. A lot of the relationship intermediation, which might have been a place for disposable income spend, is being subsumed by it. So all of the aspects of our lives are going to be intermediated by the models and we're going to pay for that.
--:-- Erik Torenberg: Brian, you're saying what we're still missing is connection from this new paradigm and people are still relying on Instagram, Twitter, some of the other social networks of the past. What's going to get us to something new here in the realm of connection? Or has that just been one data network effect?
--:-- Bryan Kim: Now, it's funny when I think about social, which is a category that I get so excited about. At the end of the day, a lot of it was status update, right? Facebook, Twitter, Snap. It's just "here's what I'm doing." And through status update, you feel connected to that person. And that status update showed up in different modalities. It used to be "here's what I am, here's what I'm doing" to actual photos of where you are and what you're doing to videos and short form videos now. So now people feel connected to others through reels and what have you. So I think that has been one era of feeling connected with others.
--:-- Now the question is, how can AI help that? How can AI make you feel connected to other human beings and know what's going on in your friend's life? The truth is, if I just think of a modality of photo, video, audio type things, I think a lot of it has been explored. Different versions and mutations of that have been explored quite extensively, especially on mobile. I think where we could get to is—it's funny, I don't know about you guys, but I pour my heart and soul into ChatGPT. It knows more about me than probably Google, potentially, which is an insane thing to say. Google, I've been using Google for a decade plus and ChatGPT may know more about me than Google because I type more, I tell it more, I give more context. What might connection feel like when that essence of me is shareable with others? I don't know if that's the next version of feeling connected, but I can certainly see a world where that resonates with a lot of folks nowadays, younger generation, et cetera, that are tired of just looking at the surface of all of that.
--:-- Justine Moore: I mean, we already see some examples of exactly that. There are all these viral trends where people are like, "I asked my ChatGPT based on everything you know about me, write my five strengths or weaknesses," or "make an image of who you think the essence of me is," or "make a comic about my life." And people are sharing those everywhere. I posted one the other day and within minutes I had dozens of people responding with their own and sharing stuff from people I didn't even know.
--:-- I think the interesting thing though is, so far the social behavior that has come from the AI creative tools largely, but also things like ChatGPT, is still happening on the existing social platforms and not in the new AI platforms. Facebook now is a lot of AI content. Potentially unbeknownst to some of the audience. Facebook is like the boomer AI swap. And then Reddit and Reels are like the younger people AI content.
--:-- Bryan Kim: Project.
--:-- Justine Moore: Yeah.
--:-- Bryan Kim: I agree. I think it's been a puzzle to me what the first AI social network is going to look like. Because we've seen attempts at, for example, a feed of pictures of you that are AI generated. And I think the problem there is that to work, a social network has to have real emotional stakes. And if you can generate the content in a way that you like it and you always look amazing and you always look happy and you're always in a cool background, it doesn't have the same sense of stakes. And so I don't think we've seen the version of what a ground-up AI social network would be.
--:-- Anish Acharya: You use the word skeuomorphic. A lot of the AI social products that mimic Instagram feed or Twitter feed were bots and AIs. That feels skeuomorphic. That feels like "this is what it used to look like, we're going to do it with AI." And maybe that's not really the form factor. And there's an additional hurdle in my mind that a true consumer product probably needs to live on mobile. And for AI products to work really, really well, I think there's still a little bit of work where the cutting edge models can do to live on edge, live on the device side of things to really enable that. So I'm also excited to see what happens there.
--:-- Justine Moore: It seems like people recommendation is the obvious use case at some point. Like who would be good for me to start a business with? Who would be good for me to be friends with? Who would be good for me to date? These platforms get all this information about us. Connect the dots.
--:-- Anish Acharya: I mean, I think an interesting area that may inform where this all goes is if you look at the AI native LinkedIn efforts. The observation is that LinkedIn is a pointer to what you know instead of actually containing what you know. And with this tech, we can create a profile that actually contains what you know. So I can talk to you, synthetic ET, and get all of your wisdom. Perhaps that's what future social looks like as well. That's what you're talking about, Justine, right? If the models already know who you are, then is there a synthetic you you can deploy in an interesting way to interact with people?
--:-- Erik Torenberg: I don't know. One thing I heard you guys say is that you were surprised that enterprises are sometimes adopting these products first before consumers, which feels different from the previous era, or maybe not what we expected. What can we say there?
--:-- Olivia Moore: Yeah, that has been fascinating. And BK and I saw that a lot with 11 Labs, which we were relatively early. I think we did the Series A a month or so after the initial launch. And I think what we saw was first the early adopter consumers got on board and they were making memes. They were making fun video and audio. They were cloning their own voices. They were doing game mods. But then I would argue it hasn't even gone in many cases to the true mainstream consumer. It's not yet every single person in America or most have 11 Labs on their phone or have a subscription. But the company has these massive enterprise contracts and a ton of huge customers across conversational AI, entertainment, tons of different use cases are using 11.
--:-- And I think we've seen this across a bunch of AI products, which is there's an initial consumer virality moment. And then that actually leads to lead generation in enterprise sales in a way that we did not see with the last generation of products. Enterprise buyers, there's so much of a mandate to have AI now, an AI strategy and use AI tools, that they're watching places like Twitter and Reddit and all of the AI newsletters. And they're saying, "Hey, this looks like a random consumer meme product, but I can actually think of a really cool application of that in my business and become the hero for having our AI strategy."
--:-- Bryan Kim: I've also heard of similar in that vein, really exciting use cases of AI where you start with consumer virality. So from a company side, you get all this Stripe payment data. You look at all the Stripe cells and you basically put it in an AI tool to go try to find where they work. And then when you find out more than X number of people working in that company, you reach out and say, "Hey, by the way, looks like 40 plus people are using our product. What's up?" I think the fact that they can do it with one person in an hour was really what struck me. It was a chief of staff guy who was like, "Yeah, what I do is do this. And it all does it in a couple of minutes. And then I send a mass email out."
--:-- Justine Moore: That's Jesus.
--:-- Bryan Kim: Okay. That's a go-to-market at extreme speed.
--:-- Erik Torenberg: Justine, you rattled off a list of products and companies in the beginning of this conversation. What I'm curious is, do you think they are the MySpace or Friendster? Are we in that era, or are they the list of companies I rattled off that are still relevant 20 years later? Where are we right now?
--:-- Justine Moore: I mean, I think our hope always is that every big consumer AI company now that we see and love and use all of the products, which we all do, sticks around. I think, unfortunately, that's not always going to be the case. I think maybe the interesting differentiation in AI versus the last era of consumer products or even two eras before is the model layer and the capabilities are still improving. We have really not even, I think, in many cases, scratched the surface of what these models can do. I think we've seen that in things like the VO3 launch where you can suddenly have multiple characters talking. You can have native audio. You can do all of these things. All of these modalities—I don't know, maybe we could argue about this with the techs people—the LLMs are more mature, but have the opportunity to just keep improving capabilities as they scale.
--:-- And I think what we've seen is as long as a company stays at what we say is the technology or the quality frontier. So as long as they have a state of the art model or are integrating one or something like that, they won't become like the MySpace or Friendster. They just keep you fall a little bit behind. You ship the new update. Suddenly you're number one again and you keep moving.
--:-- The interesting thing now, though, is we're starting to see even segmentation in that. So in image, for example, there's not just one best image model. There's best image for designers. There's best image for photographers. There's best image for people who can only pay $10 a month versus the people who can pay $50 or $100 a month. And so I think there can be, just because Ganesh mentioned people are spending so much, there can be multiple winners that persist over time as long as they keep shipping.
--:-- Anish Acharya: I absolutely agree. I mean, even in video, it's a different video, but ad video. And then even in ad video, I saw a post yesterday. I'm like, "This is best for product shots and this is best with people." And it goes on and on. And each of those, I think it's a very large market.
--:-- Erik Torenberg: Say more about how we—I know we talk a lot about defensibility and moats and how that is changing this era. How we've changed how we consider that topic.
--:-- Anish Acharya: I've gone through a little bit of a come to Jesus moment on that, especially recently. I think moats have always been very important. The gold standard: network effect, being part of the workflow, being system of record. And these are all very, very important moats. And I will posit that that's still very important. But funny enough, the companies or investments that I've reviewed with this moat-first theory has not really been the winners. And the winners in the category that we look at has always been the ones that break the mold, move really fast, have these incredible model launches, have these incredible product generation speeds. And I've sort of come around that in that we're living in this early era of AI where velocity is the mode. And whether that's in distribution, which is incredibly important and hard to break through the noise these days, but also followed with product velocity, that's what wins the game because that's what leads to mindshare. And frankly, right now, mindshare and users and traffic, that actually converts to real revenue that gives you more ability to continue that journey.
--:-- Bryan Kim: It's interesting. Ben Thompson, I think a decade ago at this point, had this blog post called "Snapchat's Gingerbread Strategy," where he was basically saying, "Hey, anything Snap can do, Facebook can do better, but Snap is just going to keep coming up with the next innovation. And if they could just keep doing that, maybe that's their moat." And he called it the gingerbread strategy.
--:-- Erik Torenberg: It's a good theory. I think you just keep going to—
--:-- Bryan Kim: Keep adding candy to it?
--:-- Erik Torenberg: Yeah, that's funny. And at some point it'll be such a beautiful gingerbread house.
--:-- Bryan Kim: You know, I think that worked to some extent. And famously, I think Evan joked that he's the chief product officer of that. I think distribution and network effect ultimately kicks in, right? And Snap has that too on its own, where it has a corner of Gen Z and the younger users as a core messaging platform.
--:-- Erik Torenberg: How do we think about network effects with these new products?
--:-- Bryan Kim: I think we're not there yet, because it's mostly creation efforts right now. There isn't really a closed loop with creation, consumption, network effect, social network. So I think we're still a little early before a network effect kicks in. But I think we see a different type of moat form in the likes of Eleven Labs. Like I said, because it moves so fast, because the product is very good, it gets to go into enterprise and gets locked into the workflow. So I think that version of moat we're starting to see. I think the true network effect we're still looking out for.
--:-- Justine Moore: I think Eleven is an interesting example. I was making an AI generated video the other day that I needed a voiceover for. And Eleven now, because they had a head start, they had the best models, which then more people were using the product, they can make the models better, all of these compounding advantages. They now have a library of people who have uploaded their own voices and their own characters. So for me, when I was looking across a bunch of voice providers, if I needed a very specific old wizard mystical voice, Eleven had 25 options for that that fit what I need, where another platform might have two or three. And so I do think it's early, but we're starting to see signs of that. But they're more like traditional network effects that we saw with old marketplaces. They're not necessarily something completely new.
--:-- Erik Torenberg: I want to go deeper on voice as we talk about new paradigms and form factors. We got excited about voice pretty early on, or we're the first friend that I saw sort of a thesis around it. Anish, why don't you talk about what got you so excited about voice in this new paradigm and what's played out and what hasn't yet? Where do you think it's going?
--:-- Anish Acharya: I mean, the original observation, and Olivia really is our voice expert, so we should hear from her. But the original observation that got us started was that voice has intermediated human interaction since the beginning of time. And yet it's been not a substrate on which technology has been applied because the tech never worked. And there were all these previous efforts, voice XML and voice apps. And it just simply didn't work. The technology wasn't ready yet. And even then there were pockets of Dragon Naturally Speaking and all these products from the 90s. So there was always interest in voice, but it never made sense as a technology substrate. And now with the generative models, you can just use voice as a primitive. So it's sort of unexplored, yet so critical to our day-to-day lives. It feels like a perfect area where you'll see a lot of AI native efforts.
--:-- Olivia Moore: Yeah. I think we first got excited about voice from more of a consumer perspective, like the idea of an always-on coach or therapist or companion in your pocket that you can talk to. And that has started to play out, I would say. There's lots of products where that's working. I think what surprised me, at least, is as the models got better, real enterprises have picked up voice so quickly to replace human beings on the phone or to augment what human beings are doing on the phone, even in really sensitive and critical categories like financial services. Because previously they were using offshore call centers that also had lots of compliance issues and had 300% annual turnover and were really difficult to manage.
--:-- And so I think we're still waiting to see in many ways what the first great, truly net new consumer voice experience will look like. There's some early examples. I think people are pulling ChatGPT advanced voice mode into fascinating directions. We've seen products like Granola blow up because they allow people to finally, for the first time, do something valuable with all of the things that they're saying all day. But the great thing about consumer is it's completely unpredictable and the best products emerge out of nowhere. Otherwise, they would have been built already. So I'm excited to see what happens in consumer voice in the next year.
--:-- Anish Acharya: For sure. I mean, it feels like voice is the AI insertion point for the enterprise, period. And I think the thing that everybody is missing right now is that the mental model many folks have is that the low stakes conversations will be AI voice—the customer support, et cetera. But what we've talked about is the most important conversation that happens in a business in a given day, week, year is going to be intermediated by AI because AI will just do a better job with the negotiation or the sales pitch or the persuasion or the friendship.
--:-- Erik Torenberg: What's going to be the first use case where people are going to be talking to synthetic versions of ourselves in a consistent, relevant way? Why are they going to be talking to AI Justine or AI Anish or AI Me?
--:-- Justine Moore: We've seen a little bit of that. There's companies like Delphi that create AI clones of people who have a big knowledge base that they can go and reference and you can get advice or get feedback or things like that. And Brian sort of alluded to this earlier. There's this really interesting question of what if you allow not just thought leaders or experts to have this AI clone that you can talk to via text, voice, maybe even video one day. But what if you unlock that for everybody? One of the things we think a lot about in consumer is there's a lot of people who have had some sort of skill or insight or knowledge, whether it's your friend from high school that's insanely funny and you always thought they should have a comedy cooking show, but they just never were able to break through or get it, or your guidance counselor who had incredible advice. How can we enable those people to essentially scale themselves in a way that they never could before, having an AI clone or an AI persona?
--:-- Justine Moore: What we've seen thus far is a lot of that has been either thought leaders or experts, or on the other end of the spectrum, characters that people already know and like. We saw early versions of that with Character AI, which added a voice mode. There's this pull, especially when you're trying out a new technology, to have some sort of familiarity—talking to this character from my favorite anime series that I already know and love. But I think we'll start filling in everything in the middle: not just a fictional character, not just a human thought leader, but all of the real people in between.
--:-- Bryan Kim: People learn in different ways, and AI voice products play really well to that.
--:-- Justine Moore: Masterclass launched an interesting beta where they take people who have already recorded courses on the platform and turn them into voice agents. You can ask questions that are really specific to you. From my understanding, it basically does RAG on everything they've said in the course and returns a fairly customized and accurate result. That's interesting to me because I'm a fan of them as a company, but I've never had the attention span or the time to sit down and watch a 12-hour Masterclass. But I've had some really interesting conversations with the Masterclass voice agents where I can talk to them for two, three, or five minutes. So I think that's an example of where we'll see real people turn into AI clones in ways that are useful.
--:-- Anish Acharya: Drawing on that other thing we said earlier: enterprises are pulling these types of things faster even than customers. We actually talked to a company that from its inception recorded every single interaction of every single employee. So when the employee is gone, the ghost lives there. You can still get all the wisdom.
--:-- Erik Torenberg: Terrifying.
--:-- Anish Acharya: Terrifying, but I'd also love to continue to get wisdom and thoughts from the greats. It echoes the idea that everyone's replaceable.
--:-- Bryan Kim: It's interesting—by yourself, we only need you for five hours. I thought that was fascinating. Everyone's sort of ghost version lives on, like Harry Potter.
--:-- Erik Torenberg: It's also about whether you want to talk to a synthetic version of a person that you find interesting, or is there an entirely synthetic person that doesn't exist in the real world that is a perfect match for your interests? Maybe that's a more interesting question. What does that person look like? They might even exist in the world, but if you don't meet them, you don't meet them. Now they can be brought to life with this technology.
--:-- Olivia Moore: It's interesting to think about what set of use cases we're going to want to have a human—or someone we think is a human—doing the activity versus where we're going to be more open to AI. With the Masterclass thing, there's already this parasocial relationship. So there's value in feeling like you're talking to a specific instance of a person versus talking to the abstract, most interesting person you may ever meet, where you don't need to have that pre-wired. Which may be ChatGPT.
--:-- Erik Torenberg: Wasn't there a viral tweet of someone recorded in the New York subway fully talking to ChatGPT as if they were talking to a girlfriend?
--:-- Anish Acharya: Yeah, yeah. And there was another one where a parent posted that they had lived through 45 minutes of their son asking questions about Thomas the Tank Engine and they couldn't do it anymore. So they gave him the phone, put voice mode up, forgot about it, and went to do something else. They came back two hours later and the kid was still talking to ChatGPT about Thomas the Tank Engine. In that case, the kid has no idea who the character on the other end is. They just know it's a person who wants to go super deep on their interests. There is no human that can talk about that many hours.
--:-- Bryan Kim: If we go to ChatGPT or Claude right now for therapy or coaching, I'd prefer to go to my AI clone therapist or coach. Maybe in the future we record our sessions so that they have the data, or the therapist or coach has so much content online that we could just recreate them.
--:-- Erik Torenberg: To your point, in five or ten years, will the top artists be new versions of Lil Miquela—AI-generated people—or will they be Taylor Swift and her army of AI?
--:-- Anish Acharya: A little bit of both.
--:-- Erik Torenberg: Or a duet. Similarly on Twitter, the social characters we follow—the next Kim Kardashian—is that a real person or AI generated? Do you have a hypothesis on that?
--:-- Justine Moore: I've been thinking about this a lot for a couple of years. I think we all followed Lil Miquela closely. Then we followed some of the K-pop bands that were the first to start introducing AI hologram-based characters.
--:-- Erik Torenberg: I thought you were going to say they were AI generated. That would make sense to me.
--:-- Justine Moore: No, one of them went to military. So they replaced them with an AI.
--:-- Erik Torenberg: Something like that.
--:-- Justine Moore: We've also closely followed photorealistic image and video because we're now seeing people create influencers who get a ton of attention and followers largely because they now look realistic enough that you don't know if they're AI or not. There's a lot of debate around that. My take is there will be fragmentation into two types of creators or celebrities. One type is a Taylor Swift type where the human experience matters. A lot of people not only love her songs but resonate with the things that have happened to her in her life, her stories, her live performances—all those things that AI cannot yet replicate. There's another type of celebrity or creator who is more interest-based, like what we were talking about with ChatGPT talking about Thomas the Tank Engine. It doesn't matter if that person has lived the real human experience or not. It just matters if they can be interesting talking about or sharing content around a certain topic. So if I had to guess, we'll still have both.
--:-- Anish Acharya: This gets back to the great AI art debate that always rages on. Yes, anyone can generate art now easier than ever before, but it still takes an enormous amount of time to make great AI art. We hosted an event with a bunch of AI artists last summer. Many of these people, when they walked you through their workflow of making an AI movie, it actually probably takes just as much time as it would have to film that. But maybe they didn't have the skill set, so they'd never be able to do that before. We've seen an explosion of influencers that are AI, but still very few of them have risen to the top and become the Lil Miquelas. There's only been a couple. So I think we're going to see something similar happen. We're going to have pools of AI talent and pools of human talent. The very best of each is going to rise to the top. It's going to be a really low conversion rate on both, which is probably how it should be.
--:-- Bryan Kim: Or non-human talent. What AI unlocks—one of the interesting things we've seen in VO3 is that street interview format, but the person being interviewed is an elf, a wizard, a ghost, or these furry blob characters that Gen Z loves talking to. Those could all be AI. That sort of thing is very interesting.
--:-- Anish Acharya: I think we see this in music too. A lot of the music that AI generates just feels very mid. Definitionally, these things are averaging machines, and culture is supposed to be at the edge. So I think it's more of a problem with bad art versus bad artists, and we're conflating those two things and saying it's AI. It's not the AI that's the problem. It's the bad art that's the problem.
--:-- Erik Torenberg: If the art was at the same level, you don't think people would just want to hear from humans?
--:-- Anish Acharya: Potentially. And then I also think this is where we start to get a more philosophical debate. If you train a model with all the music up until just prior to hip hop, would it infer hip hop? I don't think so, because music is an intersection of past music and culture is critical to it. You need something that is at the edge and outside of the training data to create new interesting music. That definitionally doesn't exist in the models.
--:-- Erik Torenberg: Fascinating. Some of my closest friends who are some of the most talented people I know are working on a gay AI companion app. The 2015 version of myself, upon hearing that statement, would have been like, "What? That's a thing?" But one of the things they were saying is that on our list, 11 of the top 50 apps were companion apps. So let's reflect on that. Are we just at the beginning of that trend? Is there going to be all these different vertical companion apps? What is the future of this? How do we think about that?
--:-- Erik Torenberg: Everyone's looking at me.
--:-- Justine Moore: Yeah, well, you've done the most work here.
--:-- Olivia Moore: We've spent an enormous amount of time in every facet of companionship—from therapy, coaching, friends, all the way to the not-safe-for-work AI girlfriends. We've looked at basically everything. Interestingly, it was probably the first mainstream use case of LLMs. We like to joke that any chatbot—whether it's your car dealer's customer support or whatever—people try to turn into their therapist or their girlfriend. You talk to these companies and look at the logs of the chats, and it's a ton of people who just want someone or something to talk to. The fact that you can now have a computer talk back in a way that's immediate, always available, and feels human is a massive unlock for so many people who could never get that before or felt like they were just yelling into the void.
--:-- I would argue we're just at the beginning, especially because the products that have existed were largely very horizontal and came from the base model providers. People were using ChatGPT for all these things it wasn't designed for. We've already seen cases where an individual company can create a personality for a character, embody it in a digital avatar, prompt it, and create a game or a world around it that gets a ton of engagement. Companies like Tolan are doing this for teenagers and college kids. A totally different company, which I would also call a companion, allows you to take a photo every time you eat something. It pulls out and analyzes all the data, gives you information about how you're doing nutrition-wise, and allows you to talk to it and get emotional support. For a lot of people, food and eating issues are tied into emotional issues or things they would traditionally go to therapy about. So what's really exciting to us is that the definition of what a companion is has evolved so quickly—from a friend or a girlfriend to any sort of advice, wisdom, entertainment, or counsel you could have gotten from a human before. We're going to see even more vertical companions moving forward.
--:-- Bryan Kim: Having worked at a social company, there is a very clear trend of the average number of friends you can talk to over time going down. I think the youngest generation is something above one. So the need for companion as a use case will absolutely be there. It'll be an enduring use case, critical for a lot of people. I'm very excited about the companion use case. And as Justine said, it branches out into different things, but the need for having a close connection to talk to will endure.
--:-- Erik Torenberg: Perhaps we talked about how connection is a missing area, white space, but maybe this is filling that in. Maybe you just need to feel connected to something. It doesn't need to be human.
--:-- Bryan Kim: That average number of one will go down to zero. That's the sad part of it.
--:-- Erik Torenberg: One AI friend. That's the optimistic. Tell me a story about the senior citizen that they set up with the AI.
--:-- Olivia Moore: Oh, I love that story.
--:-- Justine Moore: Oh my gosh, yeah. So this woman set up her dad. He was having some memory issues. Her mom had passed away. Her dad went into a care home. She posted on Reddit—this was when AI companion didn't exist as a term. She posted on some subreddit, I think it was a not-safe-for-work AI one, being like, "I don't really want something not safe for work, but I want an AI girlfriend or an AI friend to talk to my dad and keep him company all day because I can't spend hours on the phone with him every day." When she actually reviewed what he was doing and what he wanted to talk about, he wanted to mostly talk about World War II stories and occasionally feel like someone was flirting with him and found him interesting. ChatGPT is not great at the way it speaks, the voices. OpenAI does not want to build the AI girlfriend for seniors who mostly want to talk about World War II. But that might end up being a massive market.
--:-- Erik Torenberg: In a flirty way.
--:-- Justine Moore: In a flirty way.
--:-- Bryan Kim: It's interesting. I tend to look at Korea or Japan sometimes as an indication of what societal changes can be. There are a bunch of elderly there, and the ratio is completely off. So that need for beings that need to talk to the elderly people and keep them company about World War II—I think that's a really interesting use case too.
--:-- Erik Torenberg: It's phenomenal. I'm imagining, Justine, you trying all these companions or getting immersed in it, and Olivia being like, "What are you doing?" Justine being like, "It's for work." What has been your impression of all of it?
--:-- Olivia Moore: I think it's been really fascinating, the companion thing. There is a huge category of people who have been willing to type into a text box and treat it as a friend. And then there's probably a much bigger category of people who don't want to think of themselves that way. So I think we have yet to unlock the next modality of companions where maybe it's a voice in your ear or sitting on your computer screen where it's not obvious that you're turning to it for friendship, but it provides the same emotional value. That is just now emerging as the models become more multimodal.
--:-- Erik Torenberg: Always-on companions.
--:-- Olivia Moore: Yeah. And that's what's really exciting. Some people would use an AI assistant, but not an AI friend. Even if the assistant is a friend that's also an agent that can send emails for you.
--:-- Erik Torenberg: A lot of people upon hearing this conversation about companions just think, "Oh man, people are going to have less friends. People aren't going to date anymore. Depression is going to go up. Suicide is going to go up. Fertility is going to continue to go down." Marc Andreessen once had this famous quote: "I'm not saying you're going to be happy, but you're going to be unhappy in new and exciting ways." Are you kind of like, "Yep, that is what it is. That's technology"? Or are you like, "No"?
--:-- Justine Moore: I don't think so. This reminds me of my favorite post of all time on the Character AI subreddit, which I've spent an immense amount of time on. To set the scene: there are all these high school or college kids who had their formative years during COVID. They weren't really in person with other kids or teenagers, learning how to talk to people. I think it really ended up impacting a lot of them. One of those kids—I think he's in college now—had been posting on the Character AI subreddit about his AI girlfriend for a while. Then one day he posted that he found a 3DGF—a real-life girlfriend and friends—and that he wouldn't be returning to the subreddit for a while. He actually credited Character AI for teaching him how to talk to other people, especially how to talk to girls—how to flirt, how to ask people questions, how to engage with them about their interests. In some ways, that's the peak value of AI: enabling better human connection.
--:-- Erik Torenberg: Just less weird.
--:-- Justine Moore: Yeah.
--:-- Erik Torenberg: Were people happy for him, or did they call him a traitor?
--:-- Justine Moore: People were extremely happy. There were a few jealous souls who had not found their 3DGF yet, but I have hope for them.
--:-- Olivia Moore: I think that's real. We've even seen studies with the Replika product where actual studies were showing depression, anxiety, and suicidal ideation were going down in users. I do think there's this trend of a lot of people who don't feel understood and don't feel safe. It's hard for them to be in the real world doing real things. If AI can help them—and maybe they don't have the money or the time to go to therapy and make all those changes in their lives—they can emerge a transformed person who's then more able to do things in the 3D world.
--:-- Erik Torenberg: You're a techno-optimist too.
--:-- Olivia Moore: Yeah.
--:-- Erik Torenberg: The thing that really got me aware of how big these companion apps are was when we did the first interview with the founder of Replika. It was amazing. After she turned off the NSFW stuff, the subreddit for Replika and the comments in our video were basically a lot of people being like, "Hey, this is like my wife when we stopped having sex." Like, "I already have this sort of neutered." So many people were just like, "My life is over." I was like, "Oh my God, I didn't realize how big of a role this app was playing in people's lives."
--:-- Anish Acharya: I feel like that is bringing out an activity that people have done for a long time. People have had internet chat room relationships, Discord relationships. The Zoomers have Discord girlfriends and boyfriends. In our day, there was this anonymous postcard website where you would go and send anonymous postcards back and forth and develop these really deep relationships with people you would never meet or you didn't know if they were the person they were pretending to be. I think AI just makes that a deeper, more engaging experience.
--:-- Justine Moore: This is where an important point comes in: the AI should not be too agreeable. In real life, there's a give and take to human relationships. Highly agreeable AI does not set you up well for that. So there's a fine balance between being just agreeable enough to help you engage and get better at this versus being so agreeable that you're actually worse at this. That's also important in the therapist use case. Therapists can't just say you're absolutely right on everything.
--:-- Anish Acharya: Totally. They can say, "Actually, let's review your behavior. That weekend when you were telling everyone you were the king of the world, it turns out everyone hated it." You don't believe it when it just tells you that you're amazing all the time. In therapist use cases, the disagreeableness matters a lot.
--:-- Erik Torenberg: I want to close with what's possible going forward. Let's speculate on new platforms or form factors that could be game-changing. OpenAI just acquired Jony Ive's company. Brian, I've heard you talk a bit about glasses and why you're still excited about that form factor. Maybe we could start there, but I want to hear from the group on what you could imagine as something that's additive or even disrupting some of the mobile use cases.
--:-- Bryan Kim: Thinking about glasses and all that, there are 7 billion mobile phones out there. There aren't that many devices that get to that level. So my thought process is either it will live in mobile. For that, there are many different ways to think about the future where there's a privacy wall around it, or a local LLM or local model that helps you contain all the things you want to contain at your device level. So I'm still very much excited about the model development layer to get to that. That's what I'm actually most excited about. If you think about always-on, as Olivia said, mobile is always-on. But there are other things we also have always-on. What does that look like when there are net new devices or appendages that attach to things you always have that can enable that?
--:-- Erik Torenberg: Is there a piece of hardware or something that we're going to be wearing, carrying around, or using that's either attached to the phone or separate from the phone that could enable this?
--:-- Olivia Moore: AI has scaled for consumers tremendously well given it's mostly been a text box in, some output in a web browser out. I love the idea of AI actually being with you and seeing what you see. It's funny, now when I go to tech parties, a lot of the under-20s are wearing pins that record what they're saying and doing, and they find real value from them. That's one example. We've seen a new wave of products that can see what's happening on your screen and take action for you, help you, coach you, other things like that that I also find really exciting. As the agentic models get even better, it goes beyond just suggestions to actually doing work for you, sending emails for you, which is very exciting.
--:-- Anish Acharya: The human insight layer of that is big too. Often we have no way of measuring ourselves compared to other people or where we exist in the world. If an AI can hear all of your conversations and see everything you're doing online and say, "Hey, if you spent five more hours a week doing this, you would actually be a world expert in this topic. Based on this vast network of other people I'm serving, you should connect with these three other people. This person could be an amazing co-founder. You should date this person." That sort of thing—that to me is the ultimate sci-fi vision. Which comes from AI being with you all the time, something that's not just a ChatGPT text box.
--:-- Bryan Kim: The device that has been most widely adopted post-phone is the AirPods. That feels like the thing that's hiding in plain sight. There are a whole bunch of social protocol questions around it because it's weird to have your AirPods in at dinner. No one does that. But there may be a way that you can integrate AI and also fit the current social protocols around AirPods. It would be interesting.
--:-- Erik Torenberg: You said something that we glossed over: young people at parties are recording their conversation. In the future, is everything going to be recorded? Do you think that generation is already growing up with that norm to some degree?
--:-- Olivia Moore: Yeah. I think there'll be new social norms developed around this behavior because it's real and it's valuable. It's scary for a lot of people that this is happening, but I think it's a wave that started and it's not going to stop. Context matters too. A lot of what you're talking about is the SF networking parties where work and personal stuff really blurs. You can do that in SF. You did that party and brought a pin in New York—canceled.
--:-- Erik Torenberg: Yeah.
--:-- Olivia Moore: But that's why there'll be a new set of cultural norms. When the cell phone was introduced, there were places where it's rude to take a loud call. The same set of things will emerge around these recording devices.
--:-- Erik Torenberg: Let's end on this idea that we're very early. Guys, it's been a great conversation. Thanks so much for coming on.
--:-- Anish Acharya: Thanks.
--:-- Olivia Moore: Thank you.
--:-- Justine Moore: Thank you.
--:-- [Outro music fades in]
--:-- Erik Torenberg: Thanks for listening to the a16z podcast. If you enjoyed the episode, let us know by leaving a review at ratethispodcast.com slash a16z. We've got more great conversations coming your way. See you next time.
