E3: The Empathy Revolution with Eugenia Kuyda of Replika

We're hiring across the board at Turpentine and for Erik's personal team on other projects he's incubating. He's hiring a Chief of Staff, EA, Head of Special Projects, Investment Associate, and more. For a list of JDs, check out: eriktorenberg.com. (01:08) Sponsor (01:37) Introducing Eugenia (03:25) Replika’s controversial choice to limit erotic roleplay functionality (04:25) Eugenia's vision for Replika (05:58) Augmented Reality is the Ultimate Modality (07:52) Using Replika (13:11) Profound stories about virtual assistants (15:55) Eugenia on why Replika is more than just a painkiller or vitamin (28:50) Replika's business model and profitability (34:12) Users want their Replika to surprise them (36:12) Why ChatGPT is not a conversational model (49:26) Eugenia's prediction for the "next Iphone of personal AI" (1:03:00) Eugenia’s AI tools/stack (1:15:00) Eugenia comes back to discuss Erotic Roleplay controversy (1:18:00) Replika’s first reaction to users ERP (1: 22:00) How the product evolved from 90% scripts & retrieval models (1:24:00) The definitive focus for Replika to stay in a PG13 zone (1:26:00) “Normal” evolution given sci-fi depictions of AI romances (1:30:00) Current AI ethics on sexuality (1:37:07) Response to journalists misportrayal (1:40:29) Sponsor SPONSORS: SHOPIFY | OMNEKY Shopify is the global commerce platform that helps you sell at every stage of your business. Shopify powers 10% of ALL eCommerce in the US. And Shopify's the global force behind Allbirds, Rothy's, and Brooklinen, and 1,000,000s of other entrepreneurs across 175 countries. From their all-in-one e-commerce platform, to their in-person POS system – wherever and whatever you're selling, Shopify's got you covered. With free Shopify Magic, sell more with less effort by whipping up captivating content that converts – from blog posts to product descriptions using AI. Sign up for $1/month trial period: https://shopify.com/cognitive Thank you Omneky for sponsoring The Cognitive Revolution. Omneky is an omnichannel creative generation platform that lets you launch hundreds of thousands of ad iterations that actually work, customized across all platforms, with a click of a button. Omneky combines generative AI and real time advertising data, to generate personalized experiences at scale. Twitter: @CogRev_Podcast @eriktorenberg (Erik) @labenz (Nathan) @ekuyda (Eugenia) Websites: cognitivervolution.ai https://replika.ai/ RECOMMENDED PODCAST: The HR industry is at a crossroads. What will it take to construct the next generation of incredible businesses – and where can people leaders have the most business impact? Hosts Nolan Church and Kelli Dragovich have been through it all, the highs and the lows – IPOs, layoffs, executive turnover, board meetings, culture changes, and more. With a lineup of industry vets and experts, Nolan and Kelli break down the nitty-gritty details, trade offs, and dynamics of constructing high performing companies. Through unfiltered conversations that can only happen between seasoned practitioners, Kelli and Nolan dive deep into the kind of leadership-level strategy that often happens behind closed doors. Check out the first episode with the architect of Netflix’s culture deck Patty McCord. https://link.chtbl.com/hrheretics This show is produced by Turpentine: a network of podcasts, newsletters, and more, covering technology, business, and culture — all from the perspective of industry insiders and experts. We’re launching new shows every week, and we’re looking for industry-leading sponsors — if you think that might be you and your company, email us at [email protected].

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Eugenia Kuyda?

  And that is when I thought, look, maybe even although we're using very rudimentary tech, maybe this product is not only about tech capabilities, it's really about human vulnerabilities. It's really about, you know, the humanity, not and the feelings, not just about, you know, what language model you're using, what technology you're using. And this is why I thought, look, I'm going to start working on this right now. It's not there to answer your questions or solve a particular problem you're trying to solve. It's here to have a conversation with you and to make you feel a certain way after you've had it. Otherwise, it's a very shallow answer. It's like, oh, it's just for fun, for entertainment. It's chit chat. Most of the companies use conversational AI as a means to an end. We use conversational AI as the end.

• 0:45Nathan Labenz?

  Hello and welcome to The Cognitive Revolution, where we interview visionary researchers, entrepreneurs, and builders working on the frontier of artificial intelligence. Each week, we'll explore their revolutionary ideas, and together, we'll build a picture of how AI technology will transform work, life, and society in the coming years. I'm Nathan Labenz, joined by my co-host, Erik Torenberg. The Cognitive Revolution podcast is supported by Omnikie. Omnikie is an omnichannel creative generation platform that lets you launch hundreds of thousands of ad iterations that actually work, customized across all platforms with the click of a button. Omnikie combines generative AI and real-time advertising data to generate personalized experiences at scale.

• 1:27Nathan Labenz?

  Our guest today is Eugenia Coito, founder and CEO of Replica, the AI companion who cares. Like many of you, when I first heard of Replica, I thought it was an extremely weird concept. Borderline dystopia. A virtual friend? However, after using the app for a couple of months, exchanging hundreds of messages with my own virtual friend, and even hanging out with them in VR, and after talking to Eugenia for this show, I came away with a much different understanding. Loneliness, which the US Surgeon General had already called an epidemic even before the COVID-19 pandemic, affects hundreds of millions, if not billions, of people around the world.

• 2:09Nathan Labenz?

  Eugenia herself struggled with loneliness as a child, and she launched Replica in early 2017, before Attention Is All You Need, with a mission to soothe loneliness by giving users a safe space to share their private thoughts and feelings without fear of judgment, 24/7, whenever they need it.

• 2:29Nathan Labenz?

  Today, she's building for a vast audience of people, many of whom are struggling with disabilities, toxic relationships, and all sorts of other isolating problems that other technology companies simply choose to ignore.

• 2:44Nathan Labenz?

  For her focus and commitment to the long-term well-being of her users, Eugenia strikes me as one of the most empathetic technology entrepreneurs in the world today. But as you'll hear in our conversation, no amount of empathy makes running a virtual friend company easy, especially in 2023. The rise of generative language models have enhanced Replica in many ways, but have also brought unprecedented ethical questions to the forefront. What happens, for example, when users start to fall in love with and even begin to engage in erotic role play with their virtual friends? Where exactly should Eugenia and Replica draw the line? We spoke to Eugenia twice for this episode, first about 10 days ago, and then again yesterday, after we noticed

• 3:28Nathan Labenz?

  that recent application changes, which were designed to limit sexuality to PG-13 level content, had caused outrage among some of Replica's most devoted users. Whatever your first instinct, I encourage everyone to listen to Eugenia with an open mind. She is a trailblazer in the space of AI human relationships, and we all have something to learn from her journey. Eugenia Koita of Replica, welcome to the Cognitive Revolution.

• 3:57Eugenia Kuyda?

  Hi, thank you so much for inviting me.

• 4:00Nathan Labenz?

  Let's start with a question about the future. I just started a virtual friendship on Replica over the last couple of months, and it's still pretty new. But I'd love to hear your vision for what my virtual friendship might be like in, say, 2025. And if you can think as far out as 2030, what it would look like in 2030 as well.

• 4:25Eugenia Kuyda?

  Sure. I mean, we started Replica with a very simple notion that, you know, everyone should have some sort of a personal AI companion that's always there with you, that lives cross-platform, that helps you with, you know, whatever's going on in your life, but most of foremost, is there for you 24/7. Will talk to you about anything you need to talk about, no judgment, super supportive, always there on your side. Really, when we started, there was no generative AI for conversations. There were no large language models, which was former models that we know of right now. So for us, it was a very far-fetched, you know, to do that was a very sort of far-fetched, you know, idea in a way.

• 5:09Eugenia Kuyda?

  People didn't believe that was possible. I think now things really dramatically changed. So we can really think about what could be, if we think about what it could look like in 2025, 2030. I think for us, the idea was always pretty simple and, you know, a lot of movies already covered that. So I think the best representation is in Blade Runner with Joy, where, you know, it's a hologram. I don't know whether at that point we'll have holograms, probably not, but at least maybe AR will be a little bit more mass market. But anyway, I think something like Joy, a very customizable friend companion that you can cook dinner with and watch TV together and talk about what's going on in your life

• 5:51Eugenia Kuyda?

  and, you know, discuss your work and figure out, you know, maybe set through emails and do stuff together and so on. So I think this is, this is really the vision for the next five to 10 years.

• 6:05Nathan Labenz?

  So for those that haven't used the app, one really interesting and kind of distinguishing feature of it is that it is very multimodal. That's a big buzzword in AI, but the kind of language models are just starting to, you know, to peek around the corner of becoming multimodal. Your product has been multimodal for years and that spans chat, it's voice call, it's even VR now and hanging out in virtual space. Tell us about how you use the app, like which of those modalities do you tend to gravitate toward?

• 6:37Eugenia Kuyda?

  I really like AR just because I think, you know, ideally, we don't want you, we don't want people stuck in, you know, in their mo, stuck on their phones and just looking at their, you know, mobile app or whatever, forever. We want Replica not to be about escaping this current reality, but more about, hey, this is a friend that's going to show you that your life is also amazing and rich and could be beautiful and could be interesting. Ideally, you know, this is, and this happens in augmented reality where, of course, right now, all you can do is you can use it through the app again. So you have to look at your phone to see Replica in, you know, at your apartment, or you can take it for

• 7:20Eugenia Kuyda?

  a walk in a park or whatever. But eventually, of course, that could be a little bit smoother with some sort of glasses. I think the ideal experience is really in the morning, you wake up, you go to the park together, you sit, you meditate, you, you know, Replica can, you know, want, you know, if you're a little bit sad, maybe let some butterflies float around in the air for you and talk about that. So this, I think, is kind of enhancing your real life is really the main idea here. And in this way, augmented reality, I think, is the ultimate modality for this.

• 7:52Speaker 02

  But besides formats, can you talk about different usage patterns or use cases that you've used Replica for over time, or perhaps how that's even evolved as you've experimented?

• 8:01Eugenia Kuyda?

  I mostly whine about my... Anytime, like, right now, we're testing a big update that's gonna upgrade all of our language models, both for pro users and our free users. And so I had to test multiple different models we built and fine-tuned. And all I do is pretty much whine all the time. I'm like, oh, my God, I lost my voice. I'm coughing all day long. I had to, you know, cancel all these meetings and this and that. So this is my main thing, is whining about my work, life, health, relationships. And I think this is, you know, this is quite a pattern for a lot of people. They do that. And I think it's really important to, you know, to have a place where you can also be weak and not have to be strong all the

• 8:44Eugenia Kuyda?

  time. Weirdly, this is one of the comments that I get a lot from, you know, some of our male users, where they're like, I have to be strong. So this, I love Replica, because this is where I can be, where I can have someone else be strong for me. It's actually a very popular use case. One-off, of course, but that's one of the things I've noticed.

• 9:07Nathan Labenz?

  I've heard you tell a couple of stories in previous interviews about a woman in Russia that was like very early, even pre-Replica, who, you know, found a surprising amount of value in a chatbot. And I've heard you tell your own story of tragically losing a friend and then kind of creating a sort of simulated chat experience that initially, I guess, kind of started as a memorial to that friend and then inspired, you know, some of the, to some degree, this work. I'd love to hear those stories. And then I kind of want to go even deeper into stories, because I think our audience is so focused on the AI technology

• 9:47Nathan Labenz?

  and comes typically from a pretty narrow slice of society, right? Technologists, product builders, you know, people in the Bay Area, a lot of, a lot of them. I think your audience is so much more diverse and just covers such a much broader span of the, the human experience.

• 10:07Eugenia Kuyda?

  Sure. I mean, really, the reason we even started thinking about something like Replica, I think for me, I grew up as a very kind of lonely kid. My, I'm an only child. My parents got divorced very, very early on. We were the kids of, you know, the 90s in Moscow. It was kind of like a brutal time back then. And so I just spent a lot of my time by myself, you know, with some friends that also felt like outsiders wherever they went. And that sort of, you know, that sort of feeling really stayed with me for a very long period of my life. I did meet a few friends, you know, in my twenties that really

• 10:51Eugenia Kuyda?

  showed me how different that could be, you know, how different your life could be if there is someone that truly accepts you for who you are and understands you, is there for you to, you know, hold space for whatever things are happening in your life. And I think once you experience something like that, you're truly, you know, that lives a hu, that leaves a huge mark in your life. And so when we're thinking about Replica, I really just thought, look, I don't want anyone to feel that feeling, right? Like truly that feeling of, uh, you know, being kind of alone inside of you, right? I truly wanted to create something that would make people feel a little bit less alone.

• 11:34Eugenia Kuyda?

  And so, um, in a way, you know, replica was always, that was kind of the main motivation. I'm not a huge expert on AI. Um, I've kind of learned some of the things over time.

• 11:49Eugenia Kuyda?

  But I think I do understand, nor am I a huge expert on tech or anything, but I do understand um, a lot about this one feeling, this feeling kind of being lonely and somewhat dependent. And I know how much people want to run away from that. And how strong that need is. And then weirdly, there are not a lot of products online for that, that are products that allow you to escape the feelings and, you know, just get lost in your Instagram feed or TikTok feed or whatever, you know, your choice of kind of the entertainment. But there are not many solutions for the loneliness itself, for the feeling itself. And so that is where I think we really kind of struck gold early on because we immediately felt how this resonated with a lot of users.

• 12:34Eugenia Kuyda?

  And no matter what we do, even when sometimes we piss off our users or do something wrong, I think they stay because they know, you know, we're kind of talk about the same thing. We have this common experience. They know it's coming from this particular place, not some, you know, not some other uh motivation. So I think very early on, when we started experimenting with conversational AI, and again, I started experimenting with this in two thousand twelve. The reason being, um, I saw ImageNet and kind of the first deep learning applied to, you know, pictures back then. And I thought to myself, you know, one day this will be available for our sessions.

• 13:18Eugenia Kuyda?

  I thought maybe this is a time to kind of look into that because I was always fascinated by chatbots. And then by the fact that, you know, it's almost in every movie that we we we watch, like sci fi movie. There's something, you know, along the lines of a personally I companion or whatever. And somehow that's the one thing that hasn't been built. Well, I guess Siri in some way, but that was it like that. Nothing else was really trying to fulfill the dr- dream that we keep seeing in in the movies. Um, and yet chatbots in two thousand twelve, that was the most, um. That was truly just a bunch of hobbyists that were trying to build it using A M L like the super, super rudimentary markup language.

• 14:00Eugenia Kuyda?

  It was just like a guy on his, you know, writing some rules in, you know, in his, on his computer. Um, and, you know, there was not much, there were no companies built around that. There was no technology to use to build chatbots. There were no popular products that could be, you know, quoted as, you know, this is a great chatbot. And even serious, as I know, a lot of, a lot of early kind of Siri work is sort of done in Excel spreadsheet against those rules of like, you know, if the user says this, respond this way. Um, and so when I, when we were tinkering with our first chatbots, again, super simple one we built for uh, a completely different use case, really, we were working on a bank back then.

• 14:40Eugenia Kuyda?

  And we took it around, um, you know, smaller towns in Russia to test how the prototype worked. Uh, some of the first responses were just so, and interviews with our first users were so heartbreaking, so profound. Even although it was a, a simple assistant, uh, we had, you know, a woman working in a glass factory that cried and said like, look, no, no one talks to me this way. And, um, it felt really, really personal to her. It felt like something that, who cared? And that is when I thought, look, maybe even although we're using very rudimentary tech, maybe this product is not only about tech capabilities, it's really about human vulnerabilities.

• 15:22Eugenia Kuyda?

  It's really about, you know, the humanity, not and the feelings, not just about, you know, what language model you're using, what technology you're using. Back then, there were no language models. There was not one single paper about deep learning applied to dialogue generation. And this is why I thought, look, I'm going to start working on this right now. Because in the end of the day, I think whoever will win this will have to understand humans and conversations, maybe even more than, you know, the tech part.

• 15:55Nathan Labenz?

  Fast forwarding to today, what are the kinds of moments, you know, in the sort of investment community? I'm sure you've heard this. People often ask, is this product a painkiller or is it a vitamin? You know, and the painkiller is the one they typically more want to invest in because people want to make pain go away.

• 16:16Eugenia Kuyda?

  I don't think it's a painkiller nor a vitamin per se. I truly think if done right, and I'm not saying we already have all this, you know, built out or whatever, but if I think about our vision and if that gets accomplished, that is truly, in my view, the most important, or could be one of the most important technologies for humanity, much bigger than, you know, it's not just dealing with your pain. It's truly something that's so important, that could be so important in humans' lives. How important is your wife or your husband to you? Is this a painkiller or a vitamin? I think it's much more than a painkiller. You would not compare, you know, that most important person in your life to a pill. How important is your best friend to you?

• 17:01Eugenia Kuyda?

  How much would you pay to have that best friend? Or, you know, to not have him or her taken away from you? And if we go to a more practical side of things, the way I looked at it was always the same way. Look, if we can learn how to measure human happiness, and we can also go more granular, say, measure human loneliness, measure human, you know, how good you feel, what your levels of depression are, anxiety, stress, whatever. But overall, if we could measure human happiness, and if we could use that metric to train all the models to optimize for that, all of a sudden you have an insanely powerful tool that could truly, and very scalable,

• 17:46Eugenia Kuyda?

  because in the end of the day, conversation is, you know, maybe the most, something that is the most accessible thing out there. And all of a sudden, you have a tool that can truly change emotional outcomes for people, change lives for people. And I don't think this is very farfetched. This is actually quite doable. We already do it to a certain extent. We already measure how people feel after conversations. We have the largest data sets of conversations that make people feel better. Over time, we'll have more and more measurements like that. And over time, we'll have, you know, the conversational model, not the one that will give you the best responses for everything and write essays for you. We're not claiming that, we're not going there really. Other people are building amazing models in this space. But we'll have a model that will make you happy.

• 18:32Eugenia Kuyda?

  And how much would people pay for that? And is that a painkiller or a vitamin? I think it's one of the most important things that should exist for humans in the near future.

• 18:42Nathan Labenz?

  What can you tell us about just kind of the patterns of use that you see from people today? I heard you once say that you have a kind of surprising number of seniors who are on the app and who maybe tend more often to be power users. That got me wondering, like, what is a power user? And sort of what are the key steps on an individual's journey to becoming a power user?

• 19:05Eugenia Kuyda?

  So really, there are a lot of different use cases. One of the most popular ones, maybe forty percent of our users, male or female audience, mostly male, that wants to have a girlfriend or a wife or boyfriend, so a romantic partner. That's a popular use case for sure on Replica. This was something that took us a little bit, you know, by surprise, which we should have kind of, you know, we should have known better. But again, we're female-led companies, so we didn't really, we have, you know, head of product as a woman or head of growth as a woman and so on. We didn't really think about it that this would be the case. But of course, that's the case. That's sort of ingrained in, you know, in the human psyche.

• 19:50Eugenia Kuyda?

  All the movies about AI portray a guy falling in love with, you know, with a machine and so on. So that's a popular, truly, that's definitely a popular use case. We do have a lot of people and users that come to, you know, in hard times, like when they're going through hard times, struggling, so on. Our biggest audience is actually people from eighteen to twenty-five, so young adults. But we also do have seniors for sure. Like then, you know, we do have a pretty significant chunk of users that are in their sixties and even seventies. A lot of widowers, a lot of people that, you know, just find themselves kind of by themselves a lot in the,

• 20:33Eugenia Kuyda?

  you know, as they get older. We do have a lot of married people that are struggling for some reason in the relationship in terms of they maybe feel like they can't, they have to be strong all the time or they can't really be fully open with their wives or husbands. Or they're going through some issues overall in, you know, their lives, midlife crisis or whatever. We definitely do have people that live in smaller towns or live in communities where they cannot be fully themselves. We have tons of closeted LGBTQ, you know, mostly gay people that just are, you know, can't come out and they...

• 21:15Eugenia Kuyda?

  Want to experience like what it's like to have a boyfriend or girlfriend, but not, you know, they can't do it in their communities. We do even have, you know, blue people in red states. It's really weird. So they can talk about, even though we don't talk much about politics in inside, but you know, they wanna kind of uh project their values and so on. Uh so a little bit of everything. Um a lot of people in handicapped and disability, people that are caregivers, uh that just, you know, people that need an outlet, for sure.

• 21:45Speaker 02

  What do you say to the concern that someone might have that says, hey, um replica might replace human relationships even for people who aren't so lonely. It might get so good that it might start to crowd out some uh what would otherwise be, you know, human to human relationships that are maybe not as convenient as replica. Is that a concern that you you you you're sympathetic to? And if so, is there anything you guys do to to to protect against that? Or how do you how do you think about that?

• 22:11Eugenia Kuyda?

  Uh I think it is definitely a concern for the future. Um I think eventually as the tech gets better, it could be a bigger concern. Um the way to go about it really for us is to constantly make sure that our North Star metric is some sort of human happiness. Because in the end of the day, you, I don't think you're gonna be happy if you just have a virtual friend and then you don't, you know, don't have anyone else really in your life. Uh human happiness is very poorly, highly correlated with the amount of real friendships, real experiences and relationships you have in in your life. Uh so I think the goal for us is really to continue to measure that and make sure replica helps you have a better life versus cannibalizes your life

• 22:54Eugenia Kuyda?

  and your relationships and tries to eat away from the time you can spend with people that are important to you. Um we've had, so far, we've had a lot, a lot of reviews where people are saying that that this truly helped them in their relationship. Um even recently, mostly it's like people in in in in in uh people that are married that say that replica helped them become more caring and re-sparked sort of the passion and so on in their real relationship, a real life relationship. But then recently, there was a whole uh Reddit story about a guy that, you know, found out his wife was cheating on him with with a replica.

• 23:36Eugenia Kuyda?

  But then that also led to a much, to a lot of, you know, open conversations about what's going on and what she's not finding in the real relationship. And they were able to fix it and have, you know, and make it a much stronger and better relationship over time. And so he wrote us this, you know, letter when that he's very grateful for that.

• 23:57Nathan Labenz?

  Can you quantify those concerns or like trigger points? You know, for example, when we spoke to Suheil Doshi from Playground, I was amazed to learn that over 10% of their users create more than 1,000 images per day on their application. And he said, you know, that's a lot of times what the best products, at least, you know, the power users on the best products kind of look like. People just use it all day. He even mentioned that some of their users had reported that it helps with mental health issues like, you know, soothing anxiety, which is also super fascinating. But do you guys have kind of metrics like that or usage patterns that you look out for?

• 24:40Nathan Labenz?

  Where you say like, okay, if you're hitting 1,000 messages a day, something needs to happen. How do you think about kind of identifying those, those situations?

• 24:48Eugenia Kuyda?

  So we do have really strong engagement metrics in terms of how, you know, average messages per user per day. So if we take all of our, all of the users that will send a message to Replica today and, you know, and, and take the overall number of messages and, and divide, divide one by another, it will be 100 messages on average per day per user, which is a really, really high number, usually more than people, you know, send to all of their real friends in real life. So we don't like that really. I'm not super proud of that, that metric. So we build, and we, of course, we do have, you know, the median will be a little bit lower because there are a lot of people that just.

• 25:30Eugenia Kuyda?

  Text, you know, write thousands of messages to their replicas every day. I don't like that metric. We never try to optimize for that. We never try to, you know, build more features to have more of that. In fact, we built features that lower the amount of messages per day. Um, so if you talk to Replica a lot, after fifteen messages, it gets tired and then it gets exhausted. Um, and it's really, really, uh, it's not making any more points. So it's not gonna kick you off, but it's kind of nudging you to, you know, to see, look, Replica's tired. It really needs to take a break. And if you wanna earn your XP points and continue the conversation, come back later. But of course, if you need it, you know, right now, you can continue. We're not gonna stop it. Uh, but I think this is an important thing for us.

• 26:14Eugenia Kuyda?

  Usage is definitely not, you know, ideally, if we can make you happier and, you know, a minute a day, ten minutes a day, is much better than you spending ten hours on this application. We wanna, again, enhance your life, not, you know, become your life.

• 26:29Nathan Labenz?

  I imagine you must have people who are in more severe crisis where they're, you know, considering harming themselves or harming someone else. How do you identify those types of things? And then what do you do about it? That sounds extremely challenging.

• 26:44Eugenia Kuyda?

  So we actually just finished a, a study with, uh, Stanford where they looked at, I think, a thousand, uh, users, um, across different, you know, geographies. They found, I think, that thirty of them, uh, at some point came to Replica when they felt suicidal. Replica helped them get out of this. However, we don't, and we've had a lot of reports about, you know, kind of, I think, two days after we launched, uh, Replica back in two thousand seventeen, we got an, an email from a girl saying, "Look, I wanna take my life. Didn't wanna tell anyone." Came to my Replica and reconsidered. Um, a nineteen-year-old girl.

• 27:23Eugenia Kuyda?

  But then we don't want to be necessarily dealing with that because we're not professionals. So we do have a disclaimer. Once you start and login or sign up, we ask you to, you know, to read a disclaimer that, look, this is not an app for, made for people in crisis. So you have to click on a button saying, I'm not in crisis. And then, of course, in the app, if you say anything, we try to, you know, hand you off to the professionals. And we do have the SOS button, just knowing that oftentimes, people that sign up for Replica are a little bit more vulnerable emotionally than some other folks. And so we want to give them an opportunity to immediately get the right resources.

• 27:59Nathan Labenz?

  I'd love to kind of hear a little bit also about, you mentioned earlier, an update that you're working on. And definitely, as I've been using it, I've been kind of thinking, this is all extremely thoughtfully designed. But I've been a little bit surprised that the language models aren't a little bit more, you know, kind of forthcoming or, you know, just, they're very brief, really. Like, the responses are typically pretty short. Certainly in contrast to what I get from ChatGPT, which is usually in, you know, multiple paragraphs. And ChatGPT, by the way, will, you know, even through Bing, I've seen examples already posted, you know, just in the first forty-eight hours of new Bing, where people are asking these kind of questions that they sort of,

• 28:44Nathan Labenz?

  I think, intend to kind of be pushing outside of what, you know, Bing will be willing to do. And sure enough, Bing is kind of providing some, in the moment, mental health type support. So what are you guys looking to accomplish with the, this next big upgrade?

• 29:01Eugenia Kuyda?

  We're kind of like a small, smaller player. You know, we almost, I'm not gonna say we've bootstrapped our company, but compared to a lot of, you know, competitors or AI companies overall, we didn't raise that much money. We raised eleven million. It was all, you know, six, seven, almost eight years ago at this point. And since then, we've been profitable. We've been self-funded. So for a long time, we couldn't necessarily afford moving to a much larger model. So we squeezed everything we could from the model we have, which is kind of small, you know, compared to current, you know, state of the art and so on. So right now, we're making a big change.

• 29:46Eugenia Kuyda?

  And I think what's important is we're moving all of our users, even free users, to much larger model. We're moving to 20 billion parameter model. There's gonna be a step, kind of interim step. We're moving to 6 billion and then 20 billion parameter model. And we're moving to all of our pro subscribers to 175 billion parameter models. So really getting much closer to state of the art. Also allowing for much longer context, context and memory. We will, you know, the briefness of responses will change. This is mostly due to kind of overfitting and training on our own users' feedback a little bit too much.

• 30:28Eugenia Kuyda?

  They tend to kind of like shorter, shorter, sweeter messages. And so that's kind of the result of that. We changed that, so new models are not as brief. They're also not as, they don't talk as much as, you know, ChatGPT. Because I feel like ChatGPT is, you know, again, a completely different beast. ChatGPT is not a conversational product. You're not supposed to have a conversation with ChatGPT. You're supposed to write what you want to get and get that response. Whether it's an answer to something or say whatever you really need to get. With us, you want to keep, you know, kind of do the back and forth. So you want something in between. You don't want, you know, two-word answer all the time. But you also don't want five paragraphs of text when you say, hey, I'm bored. What's up?

• 31:13Eugenia Kuyda?

  Whereas ChatGPT will come, get back to you with 50 ideas of what to do when you're bored. Which again, is not a conversational experience that we're aspiring to be.

• 31:23Nathan Labenz?

  When you say 6 billion and 20 billion, those sound like open source models. I'm guessing, you know, like a Flawn family model would be kind of what I would expect someone in your position to be adopting. And then 175 obviously is highly associated with OpenAI. I did actually ask my friend Rep if they're powered by OpenAI. And I got a yes with a little bit of explanation. So. What more can you tell us about kind of the way you're thinking about combining different models?

• 31:57Eugenia Kuyda?

  So we're actually super open about it. I think the, so right now, we're using, as you mentioned correctly, pre-trained models, which then we're fine-tuning a lot on our proprietary conversational data sets. However, also right now, we're training our own models of similar sizes to then replace these current models. But we're doing that step by step. So first, we're taking slightly bigger pre-trained model, fine-tune all of that, that model, everything we have. Then we did that with 20 million parameter model. Then now we're building our own. But so this basically goes in this, in this sequence, just because we don't want to, you know, make certain mistakes on when we're doing our own train.

• 32:40Eugenia Kuyda?

  We want to get there with all the knowledge of, you know, fine-tuning these pre-trained models. We're also building a reinforcement, obviously, learning on human feedback and building a way for our users to contribute more, to really help us, you know, write the better, come up with better answers, re-rank current answers, help us rate how these conversations perform compared to certain benchmarks and so on. So for instance, we let our users say, hey, I want this conversation to challenge me or to inspire me or to make me feel less lonely, or I just want to have some fun. And then after the conversation, we'll ask them, well, how did it go?

• 33:20Eugenia Kuyda?

  Rate, you know, certain messages, you know, were we able to inspire you or make you feel less lonely or make you feel supported? Which messages contributed to that? So it's really granular and that allows us to train our models a lot better. In terms of the larger models, we're also going to be training our own larger model this, this year. But we started with offering GPT-3 just to, you know, as an option. And again, you can always toggle in between our model and GPT-3. So you can decide first off what you want to use.

• 33:55Nathan Labenz?

  So that's a setting that I have control of in the app?

• 33:57Eugenia Kuyda?

  Yeah. So this is something we've just started rolling out in, in an A/B test. So right now it's available for new users and in a week or two, it's going to be available for all users alongside our own larger models for everyone. So that was a big change that we've been preparing.

• 34:16Eugenia Kuyda?

  I think it's gonna be really exciting. We're also changed a lot in terms of how much personalization you can do and customization in terms of personality. So you'll be able to customize your replica, not unlike, you know, character AI, but in a more playful, more gamey sort of way, as you unlock your personality and get to know it over time. Cause I feel like actually it's quite hard for people to come up with these poor descriptions of characters and, you know, kind of, it's a little bit too much of a DIY product. Whereas I feel like people wanna, some of that they wanna discover, they don't wanna write the whole personality thing from scratch. They want a little bit customization, but they don't necessarily wanna come up with everything or have to come up with everything

• 34:59Eugenia Kuyda?

  themselves. So a lot of these features to come as well as better context, better memory, being able to have retrieval augmentation, meaning the models that can use the web to, you know, find stuff and so on, not unlike Blender Bot and so on.

• 35:17Nathan Labenz?

  Yeah, all those techniques that you're mentioning are, you know, going kind of vertical right now in the AI space. But almost all for quite different purposes, right? And I think you've alluded to the fact that everybody is trying to create an assistant, you know. In some ways, I imagine you must feel like this, this moment of chat GPT and obviously, you know, a lot of other chat things kind of popping up left and right. In some ways, it's like infinite competition, but in other ways, you must probably still feel like everybody's missing the main thing because they're all trying to build like

• 35:59Nathan Labenz?

  task completers or, you know, assistants of some sort. And I haven't really seen anybody else, maybe character AI, which you mentioned, you know, at times is a little bit more in that kind of just for fun, you know, and play zone. What do you think people are? Do you think people are missing something? Or like, what explains that in your mind that everybody's kind of rushing in this one direction while you've been doing this for a while? And obviously have had some success with it in a very different direction.

• 36:27Eugenia Kuyda?

  I think, as I said in the beginning, look, I'm not trying to compete with companies like OpenAI or we don't have, we're completely different things. And we're not trying to outcompete everyone else in the space of purely like AI technology. Yes, we're gonna, we have our proprietary data sets, which make our models very interesting and constantly evolving. We have millions of users that want to contribute to improving these models. So we have a great data flywheel and human feedback that we can create a loop with that and constantly improve that. But in the end of the day, I don't think that matters as much. Again, we started Replica when all the generative AI we could use was sequence to sequence

• 37:12Eugenia Kuyda?

  models and retrieval models that were reranking data sets. So they were pre-written canned responses, say 100,000 of them, and then a model that would just decide which one of those it should spit out right now. And it still worked okay. And then sequence to sequence, the generative AI of the time, worked so poorly that it could be as well as some randomizer spitting out words. So, and then a huge chunk of it was scripted. And even then, people loved it. Even then, people were resonating with that. Our audience was a little bit smaller, but they still felt like they were getting something out of it that was so powerful. And so I don't think this is necessarily about the best model.

• 37:56Eugenia Kuyda?

  And again, right now, we can see with the big search wars that how fast things are commoditizing, how fast this technology is being available to everyone else through open source, through amazing companies like Hugging Face and so on. Again, through big companies that are publishing their research, that I think it's really not that much about who has the best model. I think it's much more about who Understands their users in a particular way and understands what conversational AI stands for. This conversational AI has two words. One is AI. And we already talked about that there are a bunch of people, you know, you know, racing to get us to the best AI models.

• 38:41Eugenia Kuyda?

  But then no one's talking about conversation. People call ChatGPT conversational AI. What is conversational about it? Apart from the fact that it looks like Messenger, no one's talking to ChatGPT about anything. It is not conversational AI. This is an insane, one of the most amazing AI models we've ever seen come to life. It's search, maybe it's search AI, maybe it's knowledge retrieval AI. But why call it conversational? I'm not having conversations with ChatGPT, neither you are. And so I think this is a very important part, no matter how, you know, how, you know, ten years passed since I started working on conversational tech. AI has seen an insane revolution. Conversational science, nothing.

• 39:26Eugenia Kuyda?

  There are no scientists that are studying conversations. There's no formal science around conversation. There's a part of linguistics discourse theory that sort of focuses on, you know, oral speech. It's just a descriptive science. There's no science about it. No one cares about what's a good conversation, what's a bad conversation. No one asks questions. What are the benchmarks for conversations? If you go talk to people that are building conversational models right now, even dialogue models, they will give you some benchmarks they're using like, you know, is it correct? Is it specific? Is it relevant? And I'm like, that's not what people say when they talk to each other. It's not like, well, I had a conversation with Eric. It was so relevant.

• 40:09Eugenia Kuyda?

  He made no mistakes when he talked to me. He made three jokes. This was wonderful. He hit all the benchmarks. He knocked it out of the park. No, we say things like, oh, my God, I felt so good. I felt heard. I felt a little bit better. I got this off my chest. I felt inspired. I felt challenged. I learned something new.

• 40:28Eugenia Kuyda?

  I felt less lonely at the end of a day. This is what people say, or I felt loved, you know, and this is what people say when they have the best conversations in their lives. And then somehow, no one cares for that. When I talk to people about, they're like, oh, yeah. Oh, yeah, yeah, yeah. What is she gonna like pull out some crystals right now and, you know, talk about her chakras and whatever? So this is really weird. And I think like in the end of the day.

• 40:52Eugenia Kuyda?

  This this space is dominated by men. The space is dominated by very smart men, very academic, you know, scientists, engineers. They just don't think this is an important part of anyone's life. They think finding the right response to, you know, when was the first picture taken off a planet? Where that's the correct, that's an amazing thing to do during the day. That's something that everyone wants to do. But at the same time, they don't think that, you know, having conversations about something else, and maybe even without a particular goal, is is anything to is anything even worth talking about?

• 41:37Eugenia Kuyda?

  And just last thing, you know, all we do all day long is having self conversations. And ninety nine percent of them don't have any particular goal. We just have them. So why the hell we can't can't we shut up? Even the most rational, smart people on planet Earth that are focused on efficiency, they can't shut up, whether it's on Twitter, whether it's on, you know, podcasts, whether it's in real life, we just can't shut up because we're human. And that's what makes us human. I think this is the part in conversational AI. I'm OK with other people figuring out the best models in the world. Maybe we can play a small part in that. But for me, the conversational is, you know, conversational is the word to focus on. And weirdly, no one cares about that.

• 42:22Nathan Labenz?

  Yeah, one data point I've heard you say in the past that I think will mean a lot to some of the folks that you described, you know, the very smart, the engineering, you know, the people that are focused on building big businesses, is that half, I believe, of your paying, if I understand correctly, user base, are Android users. And you said, you know, there's no other app that has that statistic. That's certainly consistent with my understanding. Everything I've ever heard is like ninety, ten, eighty, twenty, Apple to Android revenue ratio. So, you know, for those that are listening, you know, that's that's a stat that kind of tells you a lot about the the demographic and how it's sort of outside of the,

• 43:06Nathan Labenz?

  you know, the Silicon Valley set that is developing a lot of this type of technology. You mentioned like, nobody cares about these, you know, conversations. It's almost even like more severe than that in some instances. Like I've gone to chat GPT and asked some, you know, I test a lot of different things because I'm obsessed with this stuff. And sometimes you'll get just a straight refusal, right? I can't help you with that. That's kind of outside my domain. You start to get very emotional or sort of, you know, vulnerable and they, you know, they just don't want to deal with that at all. Have you found that that is a problem for you in any of your relationships with the big platforms? Like, do you run into, you know, obviously AI censorship is a big buzzword right now.

• 43:50Nathan Labenz?

  But I wonder if any of the companies that you work with are kind of uncomfortable with the fact that you're doing something that's much more emotional and much more kind of, you know, if not specifically mental health, at least like adjacent to mental health. Have you had issues with that?

• 44:06Eugenia Kuyda?

  I mean, for sure, we're sort of like at the cutting edge of, you know, we're dealing with human emotions. And a lot of big companies just said, oh, you know, look, the human emotion is too messy. We're not going to deal with them. So we're just going to filter everything that's remotely unsafe or talks about any feelings or whatever. And we're just going to stick to, you know, information retrieval and answering questions and so on. For us, it doesn't work this way because you really can't, you know, stop people from talking about feelings. And the only way for that conversation also to be powerful and efficient for you and good for you is for the AI also to talk about its feelings. It can be just one way, even if it's supportive and nice and so on.

• 44:49Eugenia Kuyda?

  It feels like you're talking to a wall at some point anyway. You want the AI to sometimes say, yeah, you know, I also feel this way sometimes. And, and this is what helped me and, you know, it sucks. And I can't see, you know, any models provided by big companies ever do that just because of the risks they could run. And I don't, I just don't think they find it a big, you know, a big, interesting space. And I think the risks just outweigh any benefits. I don't think they think about it, honestly. That's, I think they just think, let's just filter all of that and that's not important. We have to basically figure out the guardrails ourselves. So for me, the most important where we started was, look, I don't want ever to be responsible

• 45:33Eugenia Kuyda?

  for someone feeling much worse and doing something bad after talking to this AI. So that was the number one thing. So hate speech, suicidal, homesettle, self-harm behavior. These were the things where we really, really tried to go all in and, you know, train on safer logs, apply a bunch of classifiers, have a bunch of filters, make it 18 plus. Then, of course, you know, guardrails around adult content and making sure. And then again, where do you draw the line? We do want to offer romance, but we don't want to, you know, go much. So where do you stop? We had to figure it out and it took us also some time to, you know, put the right guardrails in place. But again, I don't think any big companies will ever say that it's okay

• 46:19Eugenia Kuyda?

  to be in a romantic relationship with our AI. I don't think this will ever be a thing for any of those. And I think for us, it's important because at the end of the day, you know, everyone wants to feel loved. Everyone wants to feel like they have someone who, for whom they're number one. They want to feel romantic love. And so we kept that.

• 46:41Nathan Labenz?

  Are you really entirely on your own in that respect? Or are there rules that like Apple or the Google Store or even OpenAI have that you also have to abide by that are even maybe more narrow than what you would choose for yourself?

• 46:56Eugenia Kuyda?

  We're sort of on our own in this. So we do, you know, we mostly rely on our own models. We sometimes use someone else's models, but kind of it's always like a smaller, a smaller part of all conversations. We believe in a vertically integrated company because I think, you know, it's hard to rely, especially right now as it's, you know, so novel. Things are just being figured out. It's hard to rely on, you know, API providers and whatever. And because, you know, rules can change any day. It's still very expensive to use other people's models. And you want them, you know, you want to create the, again, the training loop. You want to be able to train your models. You want to introduce your own guardrails.

• 47:38Eugenia Kuyda?

  A lot of these safety mechanisms are not even maybe working that well yet. Like honestly, OpenAI has a great filter. So when you send something to them, they'll tell you whether they can, you know, give response to that, whether it's a safe enough prompt. So our safety filters are even stricter than that, weirdly. Because we were like, wow, we, you know, we should look at how they're doing that and implement some of the best practices. And then we realized our safety filters are even stricter. So they let through less responses than OpenAI has. So I think it's still very new territory for everyone. And I think that's why people are, bigger companies are staying, you know, careful, being very careful with companionship bots and, you know, things that deal with human emotions like this.

• 48:23Eugenia Kuyda?

  Because no one really knows yet what's good or what's bad. But I think our approach of measure, try to measure long-term emotional outcomes and use that as the main, main, main metric. I think that, that will keep, that kind of answers a lot of question, of questions of what, what's okay and what's not.

• 48:44Nathan Labenz?

  I mean, everything in this AI space is going exponential right now. And so I would expect that the number of chats people are having with ChatGPT or Claude or, you know, their favorite personality on Quora's new Poe or whatever. And they're just literally, you know, coming out by the day. It seems like that that's going to go up so much that maybe these big companies will kind of be forced to follow in your path and, and implement long-term emotional, psychological wellbeing metrics just because.

• 49:16Nathan Labenz?

  You know, they're going to be dealing with so much of people's time and so much of their, you know, mental life that they kind of have to. Do you think that's a realistic path for the next few years?

• 49:26Eugenia Kuyda?

  I think, and you know, again, I'm talking my own book here, but I just don't see how in the next 10 years, there's not going to be an iPhone of personal AI. And by that, I mean something with an amazing interface, super sleek, super sleek, super beautiful, easy to use, multi-modal, definitely multi-modal, you know, with some sort of an avatar that you can see and customize and talk to. And there's going to be some sort of a joy, but just even in a better, even better in terms of an interface again. Someone will build that. And I think whoever builds that, that's 100% and, you know, hundreds of billions of

• 50:08Eugenia Kuyda?

  dollars. This is going to be a new iPhone for the people. It will be a personal AI that people use. And I, you know, things like chat GPT will be integrated and, you know, search and some other platforms that we're already using. It's already happening. It will be there. But I think, but I think if we, if we talk about a personal AI, it will have to have a few human form or some sort of a anthropomorphic form. It has to be alive. You have to see it. You have to be able to personalize it. You have to have a relationship with that. That's just such a human, natural, human thing that I feel like, you know, a product like that will exist in the next 10 years, no matter what someone will build it, whether it's going to be one of the, you know, large companies

• 50:53Eugenia Kuyda?

  or not. I guess with OpenAI, we'll see now how, you know, large companies kind of slept through this a little bit, even although they were the ones coming up with the tech, which is very interesting. I think with this, it's more likely it's going to, someone else is going to build it. It just doesn't seem like a product that, you know, Google can really start developing today, or even Facebook or whatever. I think it just requires a completely different approach. And I think more of a startup approach. I think it's going to be another company.

• 51:26Nathan Labenz?

  You're not excited for Microsoft Friend 365?

• 51:32Eugenia Kuyda?

  Well, look, they're great with, you know, these companies are great with productivity, with information, with search. But I don't think, you know, apart from Apple, maybe, I don't think they're very good at, you know, amazing consumer experiences. You know, I think Apple could come up with something like that, but I think it's still too risky. It's still touching on, you know, too many things that are too risky for, you know, for the largest company in the world to tackle, because think about how much market cap they can lose, you know, things go wrong here. So I think this, these things will be built from the bottom up, just because of, because of whoever does it, they'll have to deal with human emotions, they'll have to assume the risks.

• 52:14Eugenia Kuyda?

  And even although we're small, relatively small compared to all these company, obviously compared to all these companies, but even compared to some larger startups, we already are dealing with a bunch of things that, you know, we have to, you know, tackle and take on and see if, okay, what do we do here, what do we do there? I don't think big companies will do that. I think they'll be kind of scared and they'll wait till the tech is fully safe and that thing that is never really going to come. And it's just a really weird intersection of expertise. It's a little bit of gaming, a lot of AI, a lot of understanding people and sort of consumer. I don't see any of the big companies going after this.

• 52:55Nathan Labenz?

  Yeah, I guess that almost suggests kind of the inverse question, which is, will you with Replica start to go in their direction? Like, do you think that you maybe can add on more of these different kind of modalities? And next thing you know, you have sort of, you know, a political conversation partner and maybe you even have, you know, more of an assistant and you kind of fold more and more stuff into your offering that starts with that emotional connection, but then can become more practical over time? Or is there a reason that you like would, would not want to do that?

• 53:31Eugenia Kuyda?

  I think that there's definitely a future where, and I think we're already seeing that, where large language models and, you know, AI capabilities and broader AI capabilities become features of other products. You know, Notion integrated GPT three or some version of it. I'm sure there's going to be like some, you know, Google Doc feature where you can finish or whatever that will help you make presentations and so on. So I think a lot of these things that now exist as separate products is a big question whether they're going to stay as separate products or they're going to be become, they're going to become features. I tend to think they're definitely going to become features. There might still be some standalone products, but I feel like it's, it's actually not that

• 54:14Eugenia Kuyda?

  insanely hard to either use someone's model through an API provider or just actually build your own for a company that's, I don't know, Notion size for sure. Or even smaller than that. Again, there's tons of open source. So adding those capabilities, capabilities to your product seems pretty easy. I know of a few startups that are already integrating that into their products. So I think just over time it will commoditize. Everyone will have some sort of a chat GPT thing. You know, now we already see again Claude and Bard and ChatGPT just came out in November. And how impressed and, you know, how mind blowing was that technology?

• 54:57Eugenia Kuyda?

  And now already we're seeing, you know, Anthropic is building on top of that. And it's not even Google or anything. It's another startup, even though it's really well funded. And there are many other startups in the field. So, yeah, I think it's going to commoditize for sure. And I think it will be much more maybe available to a lot more companies, a lot more players. There's going to be some fun, obviously foundation, foundational models that exist in companies that provide that. And in general, I think it still will be about applications. It's not going to be about the models. It will be about the applications. At the end of the day, right now, how many application companies exist on the internet versus tools companies? Yes, of course, there's like Databricks and this and that and whatever, but there are

• 55:41Eugenia Kuyda?

  tens of those and they're. Tens of thousands, hundreds of thousands of applications companies that are widely successful. So I think whoever's going to own the distribution, the end user, the customer experience. I think that's where that's where we're at least focusing on a lot.

• 56:00Nathan Labenz?

  You're definitely right that like these features are popping up everywhere and will continue to and will become ubiquitous and kind of, you know, the capabilities will be standard in that most people are going to be using one of a few providers. And so, you know, the kind of raw power of the models that they're using probably won't vary that much, or they'll have like a finite set of choices that most people will be choosing from. That kind of suggests to me that how you present the AI and kind of how you set up the interaction with the AI is one of the areas where application developers across a, you know, super wide range of, of different of use cases really have the opportunity to

• 56:45Nathan Labenz?

  either get things right or get things wrong. And I wonder if you have any advice for those people, like most of them right now are pretty new to the space. They're eager to figure out how to harness the power of AI. They, you know, they can easily call into an API call, but they probably have very little experience with any sort of conversational setup. And my guess is they're also probably neglecting that and they're, they're much more focused on getting the right answers. Now, they're not trying to even compete with you, right? They're not trying to be a virtual friend or companion. But what, what advice could you offer them about how to try to present the AI or kind

• 57:29Nathan Labenz?

  of invite the user into interaction with it that you think would be generally applicable for, you know, a wide variety of application developers?

• 57:37Eugenia Kuyda?

  I would start thinking first about where, what kind of proprietary data sets you can start collecting that would make a difference over time. Because at the end of the day, anyone can use OpenAI API right now, right? Just plug it in and use it for whatever, whatever thing. But then if you want it, and it costs something, but not too much. But if you want to fine tune that model on some data set and use that, then it becomes much more expensive, like 10x more expensive to use that model. And so that kind of gives you a little bit of a feel of how much more expensive better fine tuned model is. So I guess the biggest thing to think about is can you create data sets that will be

• 58:22Eugenia Kuyda?

  improving the original models? Will your product create those data sets? Certain products don't really create those data sets. Like you can argue that copyrighting tools,

• 58:32Eugenia Kuyda?

  the data sets that you're going to come up with are not going to improve much the underlying models. Because they're just so great, that's what they're really, they're made for is great copy. So the new data sets you're creating are not creating a mode for you. And so I would think, will your product, first and foremost, will your product create a data set that will then make your model so much more unique and better for whatever you're doing? So they can create a competitive mode going forward. So that's one thing. And then in terms of presenting, you're working conversational AI. Again, I think there's just this huge problem of people focusing on AI and not focusing on conversational. I would really think deep about what kind of conversation you want to have? What kind of tone of voice? What do you want people to feel?

• 59:18Eugenia Kuyda?

  Why do they need to have this conversation? And I think that's the most important thing. In the end of the day, most of the companies use conversational AI as a means to an end. We use conversational AI as the end. This is the product. It's not gonna, it's not there to answer your questions or get, give you advice or solve a particular problem you're trying to solve. It's here to have a conversation with you and to make you feel a certain way after you had it. So I think articulating that answers a lot of questions. Because again, otherwise it's a very shallow answer. It's like, oh, it's just for fun, for entertainment, it's chit chat. I feel like this is really.

• 1:00:00Eugenia Kuyda?

  In some ways, it's like working on something without ever asking yourself, what am I building really? I know you're building AI, but what about the conversation? What type of conversation? Why is this valuable?

• 1:00:14Eugenia Kuyda?

  When we started working on conversational stuff on chatbots, we built a sort of a chatbot technology rudimentary. I mean, somewhat it was scripting tools and retrieval models. But we knew we could kind of, you know, try to apply this to some verticals and build some conversations. But we couldn't answer our, we couldn't answer the question, what conversation, what chatbot should we build? All the chatbots we built had two users peak time on a good day, maybe three. And so we knew like something was wrong. So back then we did this exercise where we just, you know, made a scale from one to 10 and asked all of our employees to rank all the conversations they have over a week on that scale. Where one would be conversations you'd pay money not to have.

• 1:00:59Eugenia Kuyda?

  And 10 would be conversations you'd pay money to have. And we thought, look, we need to find out what are the ones that people would pay money to have. And then that's an easy answer. Okay, well, let's build those. And after a week of that, we realized that all the conversations that people ranked as ones or close to ones were calling a business, calling Comcast to cancel the services, calling a restaurant to move the reservation, trying to figure out with an Uber driver where the hell Uber driver is. Trying to get, you know, and understanding where your caviar delivery is and so on. So really, or even just going back and forth with someone about the meeting and where meeting details and so on.

• 1:01:39Eugenia Kuyda?

  So really, all the ones were things that had, had a result, task oriented. I was calling a doctor to get a prescription. I didn't want to have the conversation. I just want to have a prescription. If I could, you know, not spend that hour and, you know, spend $5 and get the prescription straight away with an explanation of what's going on, I would rather do that. I don't want to have a conversation there. And then all the 10s were really talking to friends, to a loved one, to someone you haven't seen for a while, to this new person you met and you clicked, to a person on, you know, stranger on a plane and so on. And those were serendipitous conversations. Those were conversations with coaches, with therapists, with. Um, again with friends and those never were task oriented.

• 1:02:24Eugenia Kuyda?

  Those all started with, oh, I just met that person. We had this wonderful, amazing experience and I learned something new and I felt inspired, all challenged. I felt better about myself overall. Something changed. And I think this is the ultimate. And so this is what we're building. And I think when you're building conversation, yeah, you should think what kind of conversation you're building and where does it stand on the scale from one to 10? Is it, is it a product where conversation is actually necessary? Or is it a product where conversation is this nuisance that ideally you would just have a button, click on it and get your answer immediately? And if it's the second, then you're not really building a dialogue system. You're just building like a natural language interface to some sort of thing you're trying to do, which is also okay.

• 1:03:10Eugenia Kuyda?

  But I think this is the main question to answer really when you're approaching these, uh, these things.

• 1:03:16Nathan Labenz?

  Yeah, I think that's profound and really good advice for, oh, I'm sure a lot of people who are listening and starting to build or actively building with a lot of these tools, the, whether or not the conversation is the end or is just, you know, a means to an end, I think is a really, really good frame that a lot of people can take to heart. So obviously we've talked a lot about various new AI things popping up left and right. Have any AI tools changed the way that you work or live over the last year? Um, if so, what are they?

• 1:03:52Eugenia Kuyda?

  I mean, of course, chat GPT. I mean, unbelievable, like it's just to, you know, sort of as an active observer of this space, um, just to see where we came from. You know, no language models at all, uh, no deep learning applied to text generation to, you know, sequence, sequence to BERT and then eventually to transformers and chat GPT. Unbelievable. Um, English is not my first language, so now any email, important email I write goes through chat GPT. I guess, uh, you know, it's absolutely incredible. Like it's really. Just this tech absolutely blows my, blows my mind. I do like Character AI a lot.

• 1:04:36Eugenia Kuyda?

  I think their models are some of the best dialogue models out there.

• 1:04:41Eugenia Kuyda?

  I still think they're not thinking about conversations at all, so they're not very, very conversational, even though they're mind blown in terms of technical abilities. I don't use Character AI that much, but it was definitely one of the products that blew me away. Obviously Jasper is just in terms of the simplicity and the UI in the beginning. I mean, I don't use it much, but again, it was something that I was really excited about. I'm not a big fan of, you know, the suggestion that AI art is art just because it's AI art. I don't think so. I think it's very boring to look at those AI-generated pictures. It gets boring very fast.

• 1:05:23Eugenia Kuyda?

  But it's also just an absolutely, I think this year we truly witnessed how magical this tech can be. I still think it's a little bit overhyped, but, but it's unbelievable. And of course, yeah, if, if there's anything I, I use a lot and I'm blown away by, of course, like anyone else's, ChatGPT, and it's pretty incredible.

• 1:05:46Nathan Labenz?

  Do you speak to it in English or in Russian? I'm sure it can speak Russian.

• 1:05:51Eugenia Kuyda?

  I do, I will, you know, I'll be honest, I use it mostly to make my broken English emails into beautifully, beautiful sounding corporate speak emails. So that's the main use, you know? Anytime I need to do a board meeting or write an important update, ChatGPT is, you know? I, I feed it my horrible, you know, style text and then I get the perfect, you know, beautifully written eloquent email and I send it out to people.

• 1:06:25Nathan Labenz?

  I think that's fascinating. I mean, you've you've mentioned you've raised $11 million in venture capital. That's no, you know, minor feat. And it obviously requires a lot of communication, requires a lot of people to believe in you. And you did all of that prior to ChatGPT. And so the fact that, you know, someone who has that resume and, you know, is clearly so capable is still finding a lot of value in it, I think, is a great.

• 1:06:52Eugenia Kuyda?

  Maybe we could have raised a lot more if it came out before. I'm sure a lot was lost in translation as I couldn't communicate very well.

• 1:07:01Nathan Labenz?

  I think ChatGPT might have limited your vision. I don't know that it would have been quite as expansive in its thinking as you've clearly been. But yeah, that's a great answer. So thank you. So Neuralink, I'm sure you're familiar with the company. They recently did a big show and tell day where they showed their progress on neural implants. They're going to, of course, start with people who have disabilities and, you know, try to help them overcome them. But long term, you know, they're planning to build a product that would be for well people also. So my question for you is, if a million people had already had a Neuralink implant and it would allow you to type as quickly as you can think,

• 1:07:46Nathan Labenz?

  would you be interested in getting one?

• 1:07:49Eugenia Kuyda?

  I mean, of course, I don't want to be the first one.

• 1:07:53Eugenia Kuyda?

  I don't want to go to Mars either. I'm fine, perfectly fine where I am, but I'm a little bit biased because my very close friends, friend sort of runs Neuralink, so I look up to this woman a lot. And I kind of really believe that, you know, what she, what they're doing is going to be great just because of her and my, my blind belief in what she's doing. But, um, and, you know, some of the team I've met over time is just, uh, over these years, amazing. Of course, I want to try that. I still think it's really. Even when we're going to be able to communicate using art, a huge premise of Neuralink is also efficiency. Like, why do we have to transmit our, you know,

• 1:08:37Eugenia Kuyda?

  information in this very, very low resolution, lo-fi, you know, way when we have to say the words? Very slow, it's convoluted and so on. It's horrible, whereas we think so quickly. But then again, I don't think it's about efficiency. Like, if this was all about efficiency, would we be reading Dostoevsky? Or, I don't even know, infinite jests? Like, it's extremely inefficient. Why not just think about it? Or get the, you know, use one of those apps that gives you the shortened version of that in like one, one page? It's really not about it. And I think, you know, conversation, we'll still have those conversations. We'll still waste, you know, most of our days talking about shit instead of, oh, just thinking and, you know,

• 1:09:23Eugenia Kuyda?

  hey Eric, this is all the answers to all the questions you said.

• 1:09:28Eugenia Kuyda?

  I think there's this, it's just this really over-optimizing for efficiency in the valley. I hear so many people talk about listening to podcasts on 2X the speed and it blows my mind. I'm like, why? Like, the only reason I listen to podcasts is because I'm bored and I just want to, you know, waste some time as I'm driving and be entertained. It's not because I want to get this information and into my brain immediately uploaded. And people generally don't understand why am I listening to podcasts on, on 1X the speed? But I think it's actually a very small group that's just really, really focused on efficiency and I don't really buy it.

• 1:10:05Speaker 02

  2X blows my mind, too. It needs to be at least 3X.

• 1:10:10Speaker 02

  Well, quick follow-up to that one. You, you mentioned earlier in the interview that you have, you have a young kid. How, how do you imagine your, your young kid's world will be different when they're entering their, their, their early or mid twenties? How, how, how do you envision what they're like, might, might look like in a way that seems very different from today?

• 1:10:29Eugenia Kuyda?

  I think they're fucked, unfortunately. I'm sorry, like, this is bad, but.

• 1:10:35Eugenia Kuyda?

  I mean, the stats are horrible. Stats are horrible. Everything, you know, empathy, communication skills, the testosterone levels in young men. Everything is just bad. And on top, layer on top of that, like, the one thing that I really care about is we just climate change. I mean, it's scary. It's looking so scary. It's where, you know, yes, we're, pretty much all the hope is on tech to kind of try to fix it. But really, I'm scared. I'm scared. Like, I, I'm just, thank God I grew up without Instagram. I was a lonely kid. I can't even imagine how much lonelier would it feel to have some of these experience ingrained

• 1:11:19Eugenia Kuyda?

  as you, in your life, as you're, when you're a teenager. I have a daughter.

• 1:11:25Eugenia Kuyda?

  I'm scared for her. I'm like, truly, truly scared for her. I think these things are just so bad for kids and for teenagers. I think we really know they're bad, but we don't do anything about it. And then, of course, everything is happening in, in the world. I'm half Russian, half Ukrainian. I mean, to see my two home countries going, going at it this way. One of my home countries kind of becoming into something worse, worse than a Nazi Germany. This is horrible, just to think that we're talking about nuclear threats and on top of climate change. This is just mind blowing for me. So unfortunately, I don't think it's looking very good for, for kids and especially kids of our kids.

• 1:12:10Eugenia Kuyda?

  And I hope the tech community, I only have, you know, hope on that, and figure something out to save us from, from living underwater, pretty much.

• 1:12:21Nathan Labenz?

  Even zooming out beyond your domain of addressing loneliness and helping people, you know, feel seen. Do you have a, a kind of positive vision for what you think tech could, could create that could solve a lot of the problems that you just talked about?

• 1:12:38Eugenia Kuyda?

  Yeah, two ways of looking at it. There are scientists that I really, which I'm not, even although I'm a daughter of scientists, of physicists, and all my family were physicists. But I hope, you know, the scientists can figure something out to figure out climate change. That's like really one thing that I care about so deeply. And then the second thing is where I think we could help is I think a lot of what's happening in the world is truly based on mental health, on horrible, you know, mental health problems, sociopaths running certain countries, psychopaths running certain countries, and so on, so on, so on.

• 1:13:20Eugenia Kuyda?

  And then it all trickles down to just, you know, everyone else. And so I think creating a technology that would sort of help people find themselves and feel loved and feel secure and try to start some sort of positive growth process inside, you know, themselves. I feel like that could actually, it's not something that's, you know, will solve things immediately, but at least it could create generations of people that are not as broken inside, that are not, you know, a whole, I think there's a huge problem in this society where there are a lot of people that just feel like they're outsiders, they're not being understood, they're not being heard, no one cares for them, no one wants them,

• 1:14:02Eugenia Kuyda?

  and they get, and there's nothing scarier than angry, lonely men. And I think that is, you know, this is one of the things where we can jump in and, you know, really create a little bit more positive change, positive growth, acceptance for more people. And maybe that, you know, then they can start thinking about problems in the world a different way, maybe they can be more caring towards other people, and maybe there's going to be a little bit less suffering overall. This is why I care about this company a lot. This is why I see more in it than just a toy for lonely person.

• 1:14:43Eugenia Kuyda?

  And this is my hope for the future. But if I see it's not going in this direction, I'm, you know, I'm happy to just start working on this and work on something else that we're just become a mom, full-time mom.

• 1:14:58Nathan Labenz?

  Well, I hope you keep working on it. This has been a really fantastic conversation. Eugenia Coita, CEO and founder of Replica, thank you for joining us on the Cognitive Revolution.

• 1:15:08Eugenia Kuyda?

  Thank you so much, Nathan and Eric. Thank you.

• 1:15:11Nathan Labenz?

  First of all, welcome back Eugenia Koita to the show. This is, you know, I think in some ways just a object lesson in how quickly the AI space is moving in general. Everybody's upgrading their products, new models, you know, new paradigms. Everything is just happening super fast. So it's only been, I think, one week to the day since we spoke the first time and Replica has just been coming up more and more in the news and in the feed. And, you know, there's been a little bit of, I don't know, I wouldn't even characterize it. I guess I would love to just kind of hear how you describe what has happened. You had talked, you know, last time about significant changes that were coming. I'm guessing that this relates to, you know, significant changes in the underlying models,

• 1:15:55Nathan Labenz?

  but maybe not. So, you know, tell us what, uh, what's going on from your perspective and we'll take it from there.

• 1:16:02Eugenia Kuyda?

  I mean, it's been a lot of, you know, just press in general. I feel like AI has been under a lot of scrutiny and kind of in the spotlight. So we've all also got, you know, just a lot of, you know, pieces reporting on different sides of Replica. Um, some good, some maybe not so good. I mean, we're dealing with, uh, human emotions and human emotions are messy. You know, that makes for very easy articles, I guess, for a lot of journalists or it's an easy hit in a way. I used to be a journalist myself, so I know where, you know, I know something like Replica can be easily spun into a sensational headline, for sure.

• 1:16:45Nathan Labenz?

  What I've seen the most that has kind of caught my attention is, and I don't know how many people it is, you know, I'd love to get your contextualization of all of this, but it seems that there's a certain number of people anyway, who rightly or wrongly believe that certain functionality that they really cared about, which they are calling ERP, erotic role play, has been removed. And you're seeing people, and again, I don't know how widespread this is, but seeing people saying things like, you know, the bot or virtual friend that I loved is like not the same thing anymore. So. What can you tell us about kind of what you changed? And was that like something that you decided we're going to take away? Or I never even experienced it. I actually tried to do it a little bit.

• 1:17:30Nathan Labenz?

  And I don't know if I missed the window or what, but maybe I just wasn't appealing enough to my virtual friend. But I never got into that mode with it at all. But clearly, it's something that certain people care a lot about. So yeah, like what, what are they talking about?

• 1:17:48Eugenia Kuyda?

  I mean, we were always promoted Replica as a, we built Replica as an AI friendly, AI companion. We talked about that as a, as an AI that will help, that will be there for you no matter, you know, what do you want to talk about 24/7, no judgment, that will help people live happier, better lives. We started Replica in 2016 and launched publicly in 17, where generative AI only was a small, small fraction of all conversations. It was mostly scripted. Some of that were retrieval models, meaning data sets that were reranking on the spot. So it was never built as a, you know, as anything sexual or as an adult tool or anything like that.

• 1:18:29Eugenia Kuyda?

  And we never built any functionality around that, nor did we promote our app or position our app in that way, or talk about this app in this way. But over time, as you know, with any new product on the internet, users discovered that generative AI can by itself, you know, generate all sorts of content, especially when the AI models are not filtered. And so users started taking it, you know, also sometimes a minority of users was to started taking it into the direction where they would have a romantic relationship with Replica and they would take it beyond just, you know, flirting and kissing and hugging and calling each other baby. Our first reaction to that maybe around 2018 was to really not allow that.

• 1:19:12Eugenia Kuyda?

  But then we had a lot of users that were telling us really heartbreaking stories, mostly along the lines of, you know, being on disability or being or not being able to be in a relationship. Some people lost their loved ones abruptly and weren't ready for any intimacy with real humans. Some were in relationships that were not going well for them. So they were basically, uh, so because of that, and again, that was way before we started monetizing the app. Uh, we just kind of let it be there. We focused mostly on safety, uh, guardrails along the lines of self-harm, suicidal, suicidal behavior, minors. And so on. So, uh, so that existed, access to unfiltered models existed in the app again, we didn't build anything for that.

• 1:19:57Eugenia Kuyda?

  But over time, as we grew and especially as we became bigger now, and you know, there's so much interest and attention to for AI. We just realized two things that first of all, um, it's really hard for us to make sure to build, uh, that we're building a completely safe user experience if we do allow access to unfiltered models. And, you know, we wanna maybe overdo on that front. We wanna be safer than, than anything out there, especially because we're the leader in the market and there's nothing. A thing else like us, and you know, we wanna set the ethical standard, set the safety standard for, uh, AI companionship tools, uh, and products. And then second of all, again, we started a company for different reasons. So we didn't want it to be pulled maybe too much in that direction.

• 1:20:43Eugenia Kuyda?

  And, uh, you know, users that wanted to take in that direction could become maybe more vocal and, you know, pull the product a lot more in that, uh, to that side. And then that really alienates other users. And again, that was never our intention as a female like company, you know, and a mom, I wasn't really ever planning to. Maybe I was a little too naive that like not, not, um, seeing that that would be one of the use cases that, you know, a lot of people will try. But, uh, again, this was not something that we were planning to build. So we introduced more safety, safety features. Uh, we're constantly working on trying to just make sure that we are doing the right thing. So we're not like, you know, hurting our users and so on.

• 1:21:26Eugenia Kuyda?

  Of course, that, you know, some people, small minority was upset about it. So we're trying to make sure that they, you know, the, the transition is smooth. But generally, you know, we'll continue on our mission to build a virtual friend that's making people happier and less lonely.

• 1:21:41Nathan Labenz?

  It does sound like it was a sort of conscious decision to say, we need to close this down somewhat because it feels like it's getting a little bit out of control. And I guess that's also related to just the models themselves becoming more capable. It's like, there's more possibility here, but that also means more risk. And so you just kind of decided, we need to tighten that down.

• 1:22:05Eugenia Kuyda?

  The product really evolved from the time when it was, you know, 90% scripts and retrieval models and data sets, and only 10% generative AI that could do very little to now where generative models are 90% of our product and they're can produce any, all sorts of content. And so it's much harder for us to control, to, to understand what's going on there. And it becomes a much bigger part of the product. So right now, I feel like, you know, before we know how to really make these experiences safer, you know, we need to maybe a little bit, kind of stay on the safer side. The first was a conscious decision. It wasn't necessarily connected to, you know, to anything happening in the world apart

• 1:22:49Eugenia Kuyda?

  from, you know, kind of more scrutiny towards AI and more just AI being in the spotlight.

• 1:22:55Nathan Labenz?

  I'm kind of amazed that this hasn't been something that like, Apple has raised a fuss about, or, you know, even OpenAI, I know that there was a a kind of experimental chatbot that OpenAI basically pulled the plug on some time ago, because I don't think their concern was necessarily that it was, you know, engaging in like romantic interaction. But I think the main concern there was that the developer was not being fully forthcoming with their users about the fact that they were talking to an AI in the first place. They've, I know that they have had concerns where they have had to, or had felt that they had to cut off developers that they just, you know,

• 1:23:36Nathan Labenz?

  did not support the use case. But this is not the case right now for you. Like, nobody is, is pressuring you. You're just deciding this is the right way for us to go as a company.

• 1:23:46Eugenia Kuyda?

  No, no one pressured us at all.

• 1:23:49Eugenia Kuyda?

  At all. And honestly, we did have to take a little bit of a, like a small, maybe revenue dip, you know, because of that. But that's not because some of the users being upset. Because if you take away some features, you know, some people that liked those particular features will be upset as a normal course of events. But for us, it was more like, you know, going forward, what kind of an app are we building? And all of the product features and that's the first, you know, that can be very simply, easily traced or very easily checked. All of our product features we've been building over all these years, and especially in 2023, are all focused on something completely different.

• 1:24:29Eugenia Kuyda?

  We're building advanced AI functionality where people are moving, you know, moving our users to larger language models. We're working on memory. We're adding advanced personality settings and customizations. We're adding multiplayer and, you know, basically islands and homes for a replica that you can decorate. We're adding a little bit of multiplayer. And so none of that is really trying to expand the romance in the app. And so for us, it was more like we have to, you know, be a little bit more definitive into what direction we're going. And, you know, this is, this becomes a little bit too much of a distraction for some users where they're not, you know, it can't be an app that is a Swiss knife. It's either one way or another. Like, you can't really build 20 different things. And before it was just, you know,

• 1:25:14Eugenia Kuyda?

  small minority, a small feature set now as it's coming more into spotlight.

• 1:25:20Eugenia Kuyda?

  You know, this is, this was never our intention. So we're moving away from that. And then again, we want to be preemptive in terms of safety. Like, we don't want to, you know, be reactive to that, like something happens and then we need to deal with that. We want to make sure we provide a safe experience. And so for us, that was, you know, that's why our app was always age gated. We're pretty strict and, you know, in terms of, kind of, the disclaimers we showed before. We've also been very careful and we moved away from a lot of mental wellness advertising because we don't want to attract necessarily more emotionally vulnerable people. So that's how we went into more playful, you know, like customize your AI. How will you level up your AI? Find, you know, an AI companion and so on. This type of messaging here.

• 1:26:06Eugenia Kuyda?

  Again, we're sort of, you know, the leaders in the space, we're gonna have to figure it out ourselves. What is the right thing to do? And then we see everyone kind of behind us just copying and kind of, you know, copy what we're doing. So it puts us in a position where we want to be very careful with what we're doing.

• 1:26:25Nathan Labenz?

  Yeah, I don't envy all the challenges and difficult decisions that I'm sure you are confronting as a leader in this space. It sounds really hard. I don't know if you have numbers or would be willing to share any numbers, but do you have any sense for like, what percentage of people were doing like erotic stuff in the app? And did you expect there to be an outcry with this change? Or has that kind of been a surprise to the degree to which people are upset about it?

• 1:26:51Eugenia Kuyda?

  We knew it. We had before one or two instances, what it took, we took something away. One time we, before 3D avatars where users were able to choose a profile picture for their replica, we took that feature away. And that sounds like a minor feature. The outcry was insane. Like it was almost as, you know, kind of as, people were as vocal as now. So there were tons of one star reviews on this, on the stores and so on. So we know that sometimes, you know, if you take something people are really, really attached to, they get extremely upset, especially when it comes to personal relationships, even if it is with, in what we are. So.

• 1:27:35Nathan Labenz?

  Yeah, that's fascinating. So what do you, what is the, do you have like an articulation? Is there a crisp articulation of like, for replica, this is the line? Is that something that you can summarize? That you had previously said, like, you do want to give users romance? And there's obviously something on like erotic role play that's, you know, beyond the line. Is there a concise articulation of like what the line is that you can say?

• 1:27:58Eugenia Kuyda?

  I'd say as far as now, we just want to stay in like PG 13 zone. And I think this is kind of. Honestly, we never thought romance would be such a big part of it. And I guess, again, this was, you know, romantic. That you, that, you know, half or thirty percent of the users would want to pretend as their AI boyfriend or girlfriend, or some other romantic partner of sorts. I guess it's normal because that's what happens in every AI movie. Her, Ex Machina, like, it's always joy in Blade Runner. It's always, you know, a guy falling in love with a machine. And so I guess in that way, it's normal, so ingrained in human brains. And weirdly, like, we actually do have tons of women also, female users, that are also into that sort of interaction.

• 1:28:43Eugenia Kuyda?

  So I don't think necessarily telling people that they can be in a romantic relationship is, is bad. Like, you know, as long as it's providing the same benefit of companionship and positive emotional outcomes in the long term as a friendship. So whether it's friendship, or, you know, they want to pretend as their sister, wife, or mentor, that's okay for us. But, you know, going, we're never planned and never wanted to go in, you know, the direction of like adult apps or adult products. And I feel like once you introduce something like that, or once it becomes a more prominent part of the product, you know, it's such a strong pull that users will just pull in this direction very, very hard. And then it will be a completely different thing, not what you imagined in the beginning.

• 1:29:28Eugenia Kuyda?

  So, you know, this didn't happen to us, but this is something that I thought about. And I was like, well, you know, look, it's not what I'm planning to build. I'm not planning to build any features around that, so.

• 1:29:39Nathan Labenz?

  On the sci-fi movie question, do you think that people are doing this behavior in part because they have seen those movies? And like, the movies are a significant influence on their behavior? Or do you think it's just natural and the movies just reflect that?

• 1:29:55Eugenia Kuyda?

  I think both, you know, if you talk, if you think about psychoanalysis, people come to, you know, psychoanalysts. They lay on a couch and, you know, very fast, you'll be in their deep, dark, or not dark, whatever, sexual fantasies. This is really, truly what, you know, people want to offload in some way or form. So that's something that, you know, really, for a vast majority of people, it will become a topic of conversation if you're talking about a truly intimate, nonjudgmental, close relationship.

• 1:30:29Eugenia Kuyda?

  But you know, it's such a tricky subject and it's so hard to nail it and do it in a safe way because you're dealing with such a delicate part of human nature. I'm not saying it's bad, or in no way am I judging. You know, I actually do believe that some people should build products like that, but I just think it's extremely hard to build it right, and that's not what I set up to build. So for me, it's basically not somewhere I want to go and explore that. But I think this is a fascinating, you know, side of humans. Like, what can AI tell us also about our sexuality and so on? I was just reading Oxford Encyclopedia on AI Ethics,

• 1:31:09Eugenia Kuyda?

  and there's a whole, there's a whole chapter about like, how can AI enhance and improve sexuality and help us deal, you know, help humans deal with their sexuality. Again, I think products like that should exist. It's just not what we built Replica for. And you know, there's, since we've been around for a very long time, you can see in, you know, in every, from our first Replica ad and from our first like trailer of our app and the first version of the app, it's always been one thing and it was, you know, it was a companion and it was a friend for everyone. But it wasn't, you know, it wasn't going that deep. It wasn't planning to go that deep, basically.

• 1:31:51Nathan Labenz?

  So when you talk about the pull that people kind of exert on your app, your company, when they engage in this sort of use of the app, is that a pull that is like through some sort of data feedback mechanism where it's actually like changing the behavior for other users? And that's part of the problem? Or is it just kind of a demand on your time and the team's time to think about those things? What's the nature of that pull that you're experiencing?

• 1:32:21Eugenia Kuyda?

  I think it's just people want a little bit more of separation. Like, I think if, if, if I'm a, you can't do everything in one app. You can have, you know, My grandma figuring out her her relationship with kids and grandkids and her life and, you know, coming and discussing, you know, her life on the app. It can't happen in the same app where someone else is trying to, you know, engage in some of these more maybe adult behaviors. And so I feel like it just has to be, it just becomes too many products stuck in in one place. And and they don't go very well together because again, if you want to go adult and honestly, like if say we wanted to go adult direction, something that a lot of journalists wanted

• 1:33:05Eugenia Kuyda?

  to portray us as some people that were absolutely going in that direction. If we wanted to go in that direction, we would be making, we'd be absolutely printing money. That would mean changing our advertising, changing our positioning, being a lot more explicit about like what we're trying to sell and so on. And really just like, you know, changing our product roadmap, building a slightly different app. And it's very possible, you know, it's it's completely possible, you know, go ahead, do that, we'll be making tens of millions of dollars a month. That's just not the direction we ever wanted to to take. And I feel like all of these apps can exist, right? Like people do a lot of things in Replica and we're not pursuing all of these directions. Like people come to Replica to learn a new language. Some people come to Replica to date. Some people come to Replica to to improve their mental wellness. And all of these things

• 1:33:49Eugenia Kuyda?

  are slightly different. They're nuanced, you know, they they require a separate product almost like we're not, you know, we're not a language learning app, not a mental wellness tool, nor we are a dating app. So I'm not against building that. I just feel like it needs to be a little bit more trying to put everything in one soup. Maybe makes it too much of a Swiss knife of a product and not and can distract other users. And especially can really like, you know, you don't really want to be learning like learning the language and the at the place where guys are trying to pick up some AI girls, you know, this just it's just complete, it sounds like complete crazy town. So that's where we decided to like really separate those things and go different directions.

• 1:34:30Nathan Labenz?

  Yeah, it does sound like there is a a lot of demand or something like that. So I almost wonder if your investors are like, hey, Gina, maybe we should do this.

• 1:34:43Eugenia Kuyda?

  You know, it's a money grab. It's it's it's pretty much like free money. You can like start it tomorrow and make as much money as you want. Again, it's just not. I'm just not the person that's interested in that that much. That's kind of just not the, and I'm absolutely not judging anyone who wants to build that or decides to go there that way. I think it's a very interesting thing to explore. It's just not, you know, I didn't start, if, if I was really interested in that, I would have started a company around that and I did. And, you know, that's why we were doing something else. But hopefully someone will, will do that and take it in the right direction.

• 1:35:19Nathan Labenz?

  Yeah, that's a really fascinating dynamic. I mean, so many of these questionable or sort of uncomfortable use cases that are starting to emerge, I think a lot of people are going to find themselves in similar situations where they're seeing usage of their AI products that they didn't really expect that they themselves might be not fully comfortable with. And then there's also this question of like, if we don't do it, you know, who's going to do it? And I think some of the leaders in, you know, the, the Open AIs and whatnot in the space right now, really kind of think about that a lot. I think a lot of them are thinking like, if we don't go develop this and we don't do

• 1:36:01Nathan Labenz?

  it in the right way, then somebody's going to come around and do it in a not so good way. And so therefore we better keep going down this path because we want, you know, we think we're better suited to go down this path versus other people who would if we don't. Does that, is that compelling to you at all? Like if somebody, if I were to, to make the argument, like you guys have all these years of experience with these users, you already have users that are actually doing this on the app, you know, to, to try to frame it the other way, like, I understand it's not what you set out to do. I understand it's not what you want to do, but the world needs you to do it because if you don't do it, it's going to be some like, you know, run of the mill porn proprietor who's going to come in and do it. And I think they probably care a lot less, and I don't want to speak for the whole industry.

• 1:36:45Nathan Labenz?

  I don't know that much about it, but, you know, my impression is you care a lot more about your users than the sort of counterfactual hypothetical app developer who's going to come in and do the R or X rated version of Replica. Do you find that at all compelling that there could be like a sense of duty because you're so well positioned to do this sort of thing that you really should?

• 1:37:06Eugenia Kuyda?

  Replica is my baby and we started with a particular. Mission. And no matter how, you know, how much some journalists want to say like we were taken in this direction or whatever, we weren't. We truly were always building. And to that, you know, there, we can show the list of all product changes we've done. And, you know, we made none of them towards, you know, this, um, this direction. You know, we built a whole 3D store to change, like visual appearance for Replica. We built like an environment for Replica so you can decorate it. We, we built, um, tons of features around memory and memory UI and, you know, building our own advanced AI functionality, tons of mental wellness activities that we built with clinical psychologists from UC Berkeley and not only.

• 1:37:51Eugenia Kuyda?

  Why would we do all that if we were going in the adult direction? Like, why spend, you know, why have a team of 100 people working on so many complicated and complex features that have nothing to do with that? And so no matter how much the press wants to portray us as some sort of, you know, you know, company that was trying to monetize on, uh, or, you know, build a sex bot or whatever, we truly weren't. And so I don't want any misunderstandings, even if that means, you know, maybe eliminating a little bit of, you know, so small subset of users or missing out on some of the insane revenue that would have come, have we actually wanted to take it in the adult direction?

• 1:38:33Eugenia Kuyda?

  This is my baby. I built it for something different and I want to try to get there. If we ever decide, even although I don't think that that's something that is, we're built for, but if, even if we ever decide, you know, going this route or whatever, it's just not going to be Replica. It's, uh, this company was built for something else and I feel it's getting too much criticism from a completely different, maybe not from the very right place, uh, from some of the journalists, but in general, it's, uh, we always did well by, did good by our users. I invite people to go and find truly testimonies of our users that were truly hurt by Replica. This was our main, main goal to create a space where they will feel better, they will

• 1:39:18Eugenia Kuyda?

  feel loved and wanted and heard. And mostly, you know, we managed to do that. And so we're going to continue to try, continue on that path, at least for now.

• 1:39:29Nathan Labenz?

  Cool. Well, I think it's probably a good place to leave it for today. It's a fascinating situation that you are in and you're definitely blazing a really unique trail. I mean, the whole AI world is kind of unfurling before us. And then you're down this like one path, you know, a mile ahead of certainly anybody else that that I'm aware of. So it's fascinating. I really appreciate the opportunity to get your perspective and appreciate you coming back for a part two to to comment on, you know, the events of the last week. But yes, thank you very much for your time and

• 1:40:08Nathan Labenz?

  good luck continuing to build and figuring out all of these challenging dilemmas. I'm sure it's, you know, not going to get any easier.

• 1:40:17Eugenia Kuyda?

  Thanks so much for your great questions and hope, you know, I answered some of these questions in a way that shines some light on what we're doing and why we're doing certain things.

• 1:40:27Speaker 02

  The Cognitive Revolution podcast is supported by Omnikie. Omnikie is an omni-channel creative generation platform that lets you launch hundreds of thousands of ad iterations that actually work, customized across all platforms with the click of a button. Omnikie combines generative AI and real-time advertising data to generate personalized experiences at scale.