Speaker 1: 0:16

Hi, this is Omid Ashtari.

Speaker 2: 0:18

And Natasha McElhone. Welcome to when Shall we Meet? Today's guest is Tucha Bullut. She recently launched Luna, a Gen AI dating coach that aims to transform the dating market. She is committed to providing a solution to what she believes is a loneliness crisis.

Speaker 1: 0:38

Previously, tucha was a strategy consultant at EY Parthenon, advising private equity houses on M&A. Tucci is a published author and a Cambridge economist specializing in the use of technology to bolster economic development. She's passionate about the power of AI and the positive change it can bring to the world. As a serial entrepreneur, tucci started a number of AI businesses, including Street Bees, where she served as a CEO for eight years.

Speaker 2: 1:06

Today we discuss whether tick boxes are really the way to find your life partner.

Speaker 1: 1:12

Has dating become yet another job?

Speaker 2: 1:15

The rapid adoption of dating apps as a way to meet your romantic partner.

Speaker 1: 1:20

How many more men than women there are in these heteronormative apps?

Speaker 2: 1:24

And how having fewer women perhaps creates some unique dynamics.

Speaker 1: 1:30

Democratizing the access to relationship coaches.

Speaker 2: 1:33

The tradition of arranged marriages a good or a bad thing?

Speaker 1: 1:37

How responsible Hollywood is for creating unrealistic expectations.

Speaker 2: 1:41

Or Bollywood or Nollywood.

Speaker 1: 1:43

Fair.

Speaker 2: 1:44

So are you ready for our date?

Speaker 1: 1:46

let's do it. Hello everyone. This is omida shtari and natasha macklehone, and with us we have today tucha bullets hi tucha, how are you?

Speaker 3: 2:01

very.

Speaker 1: 2:03

I am excited about what you've launched, and it launched very recently, so why don't you tell us a little bit about what happened last week?

Speaker 3: 2:12

Happy to. We made the cardinal mistake of launching the new solution on a Friday. Anyone in tech would know never release a new solution on a.

Speaker 3: 2:24

Friday Bad idea and we did a soft launch where we announced Luna to the world, that we are launching a Gen AI dating coach who can actually take over your dating life for you. It was a soft launch, it was a single LinkedIn post, basically, and we were expecting maybe 100, 200 signups as our beta customers. Our servers completely blew up. We got 5,600 signups over the next two days. Obviously, we spent the entire Friday night scaling up our servers. We spent the whole Saturday making sure everyone is getting through the signup process smoothly. It was tough, but we are very happy about the interest.

Speaker 1: 3:13

That's awesome. I mean, this is a good starting point for us to acknowledge that there is a moment here where a lot of people, I think, are somewhat disillusioned with, maybe, the current availability of options in the dating app space, and we will return to that topic. But maybe let's anchor by starting at the very top. Let's talk a little bit about the general state of relationship stats maybe, and how do people meet each other nowadays. I think you have some stats here, I have some stats here. Do you want to maybe talk a little bit about what your background research has been in this space when you were trying to kind of assess whether this is something you should be going after or not?

Speaker 2: 3:52

Can I also ask one question, which may seem very naive, but for listeners that perhaps have never been on a dating app when do we date this back to that dating apps became something that was used by the majority of people looking for a partner?

Speaker 3: 4:09

That's a great question. So there's a first generation of dating apps OkCupid, eharmony and they've been around for nearly 20 years, so it's not something new. 20 years so it's not something new. We then got Tinder. They discovered the beauty of swiping and we had an entire generation of dating apps that followed the Tinder methodology, who ended up dominating the space, and what we are seeing now, with the generative AI becoming available and cost effective, an entire new generation is going to emerge, which now personalizes the whole dating experience for you.

Speaker 1: 4:55

Cool. So what percentage of people do actually find each other through dating apps nowadays in the Western world, I guess?

Speaker 3: 5:01

Yeah, it's a great question. It's actually as high as 65 to 70 percent, depending on the country that you're looking at, and is this heterosexual partnerships or is that across the board?

Speaker 2: 5:12

is that average for for gay, queer, everything?

Speaker 3: 5:16

this is actually across the board. It's not that different when we look at the same sex usage. Obviously, the purpose of the using dating apps is slightly different for heterosexual versus the same-sex couples, but we still do see people getting into a relationship. What percentage of them found their partner on a dating app? It's more or less within the same region yeah, so I'll start.

Speaker 1: 5:41

similar to what you're stating here More like 50% of the couples finding themselves this is heterosexual and 65% gay and queer community.

Speaker 2: 5:50

So it's more, so there's more. There's an over-representation A little bit more, it seems to me.

Speaker 1: 5:54

Just maybe as a public service announcement. I don't have that much experience and we're going to try to obviously be inclusive here by trying to look at the stats across the spectrum and we've done some research obviously to be able to represent everybody here. But my own experience is that of a straight man and that's why I'm going to a lot of the things that I may say will come from that angle. But it does seem like the heterosexual couples are a little bit less likely to match as compared to homosexual and, again, queer.

Speaker 2: 6:29

So is that the upside of the apps is that it can create a safer space or a more specific space for what you're looking for, whereas going up to someone in real life and one first of all seeing if they have the same sexual persuasion as you, is a risk, I guess.

Speaker 1: 6:45

Especially depending on where you live.

Speaker 2: 6:47

Yeah, and so the other advantage is exactly that, if you're in a smaller, maybe rural, community with perhaps less diversity I'm not bashing rural communities in any way, shape or form, but there's just less people.

Speaker 1: 7:00

Yes.

Speaker 2: 7:00

So perhaps there's less choice of people who have your particular sexual identity. So that's opened that up for a lot of people in a way that for it's harder.

Speaker 1: 7:11

Yeah, absolutely. What about divorce rates? How are we looking on that front? Do you have any stats available on that? Have you looked at that?

Speaker 3: 7:18

Well, I think one maybe to start with is the most surprising one. Do you have any idea what's the divorce rates in India?

Speaker 1: 7:26

Yes, I do.

Speaker 3: 7:30

Shall we share.

Speaker 1: 7:31

Yeah, it's about 2%.

Speaker 3: 7:33

This did come as a surprise to me. It's something I actually found out today and it was not my expectation. I lived in India for a year before, spent a lot of time there and I understand the marriage culture is very different different. But when we look at the rest of the world before maybe we go into the divorce rates. The average length of a marriage in the us is eight years that already tells you a lot about you know, the divorce rate.

Speaker 3: 8:05

So it's quite interesting to compare us and europe, which has a much more similar rate.

Speaker 1: 8:11

Middle east even has a similar rate versus india at one, two percent what you're saying is obviously right, um, what I think is when you see that something is not only an order of magnitude but a factor 30 lower than anywhere else, there's multiple factors at play here, it's not just one right. And I think we can probably say that there is a bunch of these, um, arranged marriages and I don't like to romanticize that concept because there's obviously some of these marriages are under duress and maybe not happy and against their will and all that. But I think when there's a whole culture that actually really supports the notion that marriage is not love but it is a commitment to somebody else, without the romantic notion of having to find the perfect partner but, like you have, you know, somebody that is compatible in some shape or form, maybe that changes dynamics.

Speaker 2: 9:01

I think to add to that from my again very anecdotal, limited experience from my friends who, two of whom have had arranged marriages and are from Indian heritage, what I've noticed when I've questioned them is that it's not two individuals who are coming together, it's two families two tribes, two.

Speaker 2: 9:21

You are bringing together much more than two individuals and you have the support of that family. There was someone when I asked him you are now living in the States and you've been educated in America and you've dated women. What's made you come back home and let your parents arrange your marriage? And he said well, I trust the experience and wisdom of two generations, of two lots of families around. What might make me happy and what might make her happy? They have our interests at heart. Then my knee-jerk response to someone's sex appeal or chat in a date. I thought, god, that is culturally so different.

Speaker 1: 10:05

Yes.

Speaker 2: 10:05

Where we dismiss very often previous generations and their attitudes and think that they're out of date and that we just don't respect the wisdom. Because, really, how much of human relationships, love, dynamics, romance, how much has it changed to be loved and to be wanted and to be accepted and not rejected?

Speaker 1: 10:24

It's been around for a couple of thousand years.

Speaker 3: 10:27

There's actually probably what. The purpose of marriage is also a very relevant question. It was quite interesting for me during the years I lived in other countries, including India, that not everyone expects the same thing from a marriage. The definition of what marriage does for you is also very different. If you are going in for the sexual appeal, if you are going for feeling loved, safe, or you are getting married almost like building a company with a co-founder and we are going to take the life admin together and our job is to raise, would be the rare reason to take you to a divorce.

Speaker 1: 11:32

Let's come back to our own cultural circles. What are the divorce rates around here?

Speaker 3: 11:36

So we are looking at north of 50% in the US and even in Europe and in Japan you are looking at north of 30%, similar to Middle East. When we go to Turkey, which is my home country, where I grew up, we see also about 35% or so divorce rates. That's, of course, lower than UK, which is more around 42%, and it's quite interesting to see that there is a well-known correlation between the female participation in workforce and the divorce rate In most countries. There are definitely places like India where that trend, that correlation, doesn't hold, but it's quite interesting to see female participation in workforce in the UK is around 58%. That number comes down to 35% for Turkey, so there's clearly a relationship between the two.

Speaker 2: 12:31

So the economic dependence is a factor in not defaulting.

Speaker 3: 12:35

It definitely makes it easier to walk out.

Speaker 2: 12:37

Yeah.

Speaker 1: 12:38

Well, that's one reason. Let's talk a little bit about Hollywood's influence here as well, because I do think that you were alluding to it as we were talking about the Indian marriage paradigm. There are other expectations here, and I think a lot of the issues that we see are also unrealistic expectations as to what a relationship should look like, because the framework that one has is you know all these romantic comedies where after two weeks it's like the love of life fulfills you. Everything's amazing, everything's butterflies, and when those butterflies really go away in a real relationship, in a real life, and shit hits the fan, then obviously people feel more inclined to just quit because they're looking for that butterfly thing again to play devil's advocate, I would say bollywood was all about the happy ending and the romance and the fantastical disunification of love true, but then the the exit divorce side is not at all as available as it is in other societies I think to some extent that really comes back again to what you are expecting out of a marriage what.

Speaker 3: 13:41

What is butterflies? Is it trauma bond? Is it actually someone making you really anxious and you are associating that inner anxiety with a familiar feeling that you had when you were a little child and how our mothers primarily, but our parents in general, bonded with us, becomes the definition of love and we seek to recreate it, reproduce it. So it's actually perfectly possible to have you know, to feel very excited about someone and keep that excitement in the long term. But that excitement in the first place wasn't the butterflies, because a lot of butterflies actually come from the fact that does this person like me? There are a lot of question marks, can I win them? That whole mentality of I need to conquer people, I need to win, doesn't necessarily yield good results in the long term in the long term.

Speaker 1: 14:44

It's funny. I had an interview about music and what gets people excited and where dopamine really flows in music is when one anticipates something and it is fulfilled, or when someone is surprised. And I would say probably the same is true in relationships where the dopamine flows because when it anticipates oh yes, I'm going to be together with this person first kiss, first time, sex, first time together oh, she moves in. And then all of a sudden, you know, together with this person, first kiss, first time sex, first time together oh, she moves in. And then all of a sudden, you know it's kind of like, okay, yeah, there's not much more to anticipate here. All of a sudden, excitement goes away. Um, we, we should maybe talk, dig in a little bit into the area of dating apps in particular. Now, right, and uh, that's maybe so 20 years.

Speaker 2: 15:21

We've established roughly that this has been going on for right and everyone would accept that there was a certain stigma attached to having to join. I was speaking to a Gen Z person this morning who said, oh, that hasn't gone away, there still is a stigma. And he said, yeah, absolutely, in a student community, you should just be fraternizing with your peers. You shouldn't need to go onto an app. That's kind of the perception. And then once you leave tertiary education, if you're in it and you enter a workplace, then perhaps, yeah, you're working with the same people all day, every day. It's okay, it's considered like it's more normal. I don't know what the stats are around all of that stuff and how many very young people join dating apps.

Speaker 3: 16:02

I think this is actually a very controversial issue we need to call out and publicly discuss. How much time did you spend to train yourself to get a job in the space you wanted to go into? How long was your education to get there?

Speaker 1: 16:21

Yeah, probably more than 10 years or five years, five, six years At least specialized right.

Speaker 3: 16:27

Exactly. You know I did some of my education in Turkey, some of it in the UK, I think. I studied including my PhD about eight years learning a subject to do a job that I would really enjoy. The whole idea that you're going to run into someone in a pub and you are going to find the ultimate compatibility and love and you will live happily ever after is like less likely than winning the lottery. So if that's what you want for your odds to get a successful relationship, go ahead.

Speaker 2: 17:08

Where did you get that statistic from?

Speaker 3: 17:12

I love the question, but I think divorce rates speak for themselves. We keep marrying people with whom we don't end up having a successful relationship with, and I'm not even including here the stats of people who are in unhappy relationships are just not getting divorced.

Speaker 2: 17:31

Something else we should establish in this conversation is this very kind of heteronormative idea of marriage and you've mentioned children a few times or building a company and the kids are the endeavor.

Speaker 2: 17:43

One of the things that's happening more recently or more openly is that people are able to not sign up for that very conventional social construct or, we mentioned earlier on, they're not meeting in churches.

Speaker 2: 17:57

You know, maybe religion has taken some of the pressure as other social constructs have broken down and that sounds negative are perhaps changing and shifting and I want to get to AI because I think that's having a massive effect on how we organize ourselves. But if you look at, if you do want children, that we have children, that we leave it till later, that we employ science to help us procreate, and then an awful lot of people and a lot of my straight female friends have elected by choice not to have kids and it's become socially acceptable in a way that it just wasn't before. People have been quite declarative about that right. They'll put on their profile want kids or not interested in having kids, or you can see those things very quickly. It surprises me in an evidence-based way that, given that you can be really open before you've even met someone about what it is that you want in life. It's so dissatisfying for people on the whole likely than a lottery win or not.

Speaker 1: 18:54

But ultimately what she's saying I agree with a lot, and that is we don't know what we want. What we want is usually not what's good for us and we are not taught any of that. And we find somebody in a bar and we end up being 20. And then we kind of fall into a relationship and we think that's the reality and then in the most people get divorced in their 40s right or break up in their 40s and all of a sudden have a midlife crisis, but because they realize they lived a lie or they lived in a situation that wasn't really the right fit, or it wasn't, or it wasn't, or it wasn't because they've changed and the other person didn't want to change with them, or they weren't proper, a proper match for a life, for a lifetime.

Speaker 1: 19:44

And you know, you okay, one person can have three chapters in their life. They don't have to have a match for the lifetime.

Speaker 2: 19:50

That's the Esther Perel idea, isn't it?

Speaker 1: 19:51

And that's absolutely okay, right, I think it's just if what you thrive in doesn't match what you're pursuing, then that's a problem, right? And I think that's basically where Tudç's point comes from. If you really want to have a life with one person, then assuming that you're going to meet them randomly in a bar on a night out is probably not necessarily the highest likelihood.

Speaker 3: 20:13

I think the key thing here is democratization of access to that support and help that we are seeing with generative AI. People sometimes know that they are not sure what they want and if a very, very small group of people have access to people to discuss that with whether that's a qualified therapist or maybe like real educated parents, whatever it is, I think what we all want to do in the age of generative AI, bring that kind of support with technology at a very affordable price to the mass market, because understanding yourself is the first step of a successful, healthy relationship and, once you are in the relationship, understanding the other person and learning how to communicate, developing the muscle of intimacy which we desperately need at this stage. Because there are two sides of the coin with generative AI. On the one hand, we are seeing stats, especially from Japan, but the rest of the world tends to follow right. We are seeing from Japan.

Speaker 3: 21:25

We have all these apps like Character AI Replica Japan. We have all these apps like Character AI Replica. Virgin AI is coming in and replacing the human connection and we are seeing very interesting feedback from the users saying that I don't want a girlfriend who wants expensive dinners, I don't earn enough for that, or I don't want a girlfriend who can reject me. I don't want to deal with the feeling of rejection, so I'm opting for an artificial girlfriend who is always nice, always kind, submissive maybe, and that's exactly what I'm looking for, but we are actually completely missing out on the real human connection. So instead we can take the same technology but use it for people to understand why rejection bothers you so much.

Speaker 2: 22:10

That was going to be my next question. So we're becoming pain avoidant Pain avoidant.

Speaker 3: 22:14

We don't like being inconvenienced.

Speaker 2: 22:16

We don't like feeling uncomfortable.

Speaker 3: 22:18

A hundred percent. The reality is a full human experience in this world is going to get significantly enriched if we continue to develop that intimacy muscle. But if we actually take risks as well.

Speaker 2: 22:34

Take risks isn't. Is that part of it?

Speaker 3: 22:36

take risk is a big part of it. Do we really want to live in a world where people are just doing all their work from home on a computer in the evening, you know, taking their like super cool tech classes and living with their artificial girlfriend? Most of us do not want to live in that world, including the people, by the way, who are exercising that. They are still suffering from the lack of oxytocin, from the lack of like, a real human connection. So, coming back to your question, I was traveling in Mexico last summer and in a very small Mexican town where notaries go, we met this beautiful trans person and it was a very hard experience for them to go through that because it's a small town, you know, it's not very open minded can't find a partner. We're open-minded, can't find a partner. I think what we are really looking for with these technologies is that help everyone find their own crowd in their own vibe, whatever that is.

Speaker 2: 23:39

Nice yeah.

Speaker 1: 23:40

Yeah, whether you want to have a lifelong relationship or polyamory or whatever floats your boat really. I'll take us back into the territory of maybe looking at the stats of the dating apps right now, because I think the the current paradigm or the current avenue, where we said more than 60 percent of people in the west are meeting each other on, is it's a very dire place. It seems like that's why you had 5 000 signups for l. So let's talk a little bit about this dire space and again, this is probably from a heterosexual perspective, because it may be a safe space for others, where it's like a toxic wasteland for some.

Speaker 2: 24:21

For sure Is it advertised for straight couples.

Speaker 3: 24:25

At the moment it is. We strongly believe the match patterns and even the tone of voice that Luna should have will be different for different communities. So we are starting mainstream, we are starting with London, we are starting with heterosexual couples and what we are doing, though, in the background, we are training our own large language model, which is a domain specific frontier model, and we completely appreciate that a lot of people who are homegrown, coming from within those communities, will be the best people to train the tone of voice, so much better than us. They may not have the resources to build the tech, so we are very much looking forward to building partnerships over APIs. If a community in Mexico wants to build a version of Luna just for their own trans community, then we can help with the technology they can do. What should the app sound like, rather than us trying to find the right vocabulary, right, you know tone of voice for them?

Speaker 1: 25:34

Nice. Let's go back to the stats. So tell me, what are we seeing in terms of female male ratio on these dating apps right now?

Speaker 3: 25:44

It's a bloodshed. Dating apps are a bloodshed. It's a bloodshed. Dating apps are a bloodshed. Why? Because it's a very simple mathematical distribution problem. 75% of dating app users on average across the apps are men. Wow, that goes to 76% for Bumble. It's about 75% for Tinder, 68% for Hinge Was it always like that Not always.

Speaker 3: 26:14

I think they tend to start more 60% men, 40% women, region and then, over time, women who use the app churn faster and the word of mouth goes around that it just doesn't work for women and they stop signing up. Not only there are significantly more men on dating apps, men also tend to give more likes, so it's exponentially getting unequal between the two groups, which is leading to a place where the top 20% of men in a dating app cleans up, gets 80% of the likes. The remaining 80% of men struggle to get one date per month. And what was really surprising to me, that on average I'm not talking about heavy users, right On average people are spending 30 hours per month, that's one hour every day. People are apparently putting in their calendar, you know, after dinner to go through their dating app. It has become a job.

Speaker 1: 27:30

Yeah, it's a chore and I think what you're saying is quite interesting. So we're saying that the majority are men. Men are swiping right on every second kind of profile. Women do so on 15% of the profiles. So women are much more selective when it comes to wanting to match with a guy and there are fewer of them.

Speaker 1: 27:51

And so the experience again we're talking in a heterosexual setting of women is that they get inundated with a lot of matches and again, of course, the and I hate reducing people to this, but, like the 20 percent most successful women essentially get a lot more likes than, say, the bottom 20%. Right, there is a distribution thing even though it's not as extremist with men where, as you say, the top 20% get 80% of the likes. So we have a situation where men are just spending a lot of time trying to find someone, and essentially these platforms love that because if they want to stand out, they have to buy things that allow them to stand out, like roses, like super subscriptions, where their profiles have a golden border or whatever the else it is, so that they can have the illusion that somehow, with this extra stuff, they can be dateable by the few women of which were very selective to match with them.

Speaker 2: 28:58

Just um, to elucidate yes, with these extra things that they can buy, what do they then do? Do they? Does the algorithm then pick them up and then put them to the front of the queue or make them more visible on the app? Is that?

Speaker 3: 29:11

Yeah, that's correct. It gives you impressions. You are buying impressions effectively.

Speaker 2: 29:17

So you're buying eyeballs.

Speaker 3: 29:18

You are buying eyeballs so more people see you, or if you really liked someone, they make sure that person sees you at the top of their queue.

Speaker 3: 29:28

that person sees you at the top of their queue, but it doesn't do anything in terms of probability of that person actually liking you. That's why there's a lawsuit in the US right now resembling Match as a business who own Tinder and Hinge to a slot machine, and they are saying that they should be regulated in exactly the same way betting and gambling companies are being regulated. Most of the time you are buying roses or spotlights for people who haven't been active for maybe three months, you're still seeing their profiles.

Speaker 2: 30:07

And they have that information. Obviously, and the company has that information. Obviously the company has that information.

Speaker 3: 30:12

The user doesn't have that information. It's a very opaque sector, so they are showing you incredibly appealing profiles who have not been active or have so many likes already. For them to get through all those likes and get to you is highly unlikely. The reality is any matchmaking company has the stats on probability of two people actually matching. So if you did use that information, you can significantly decrease the number of people you have to see on the app. But the problem is the goal of the user on the app does not align with the business model that these companies today have. The more roses you send to people who are actually never going to match with you, the more money they make. The longer you stay single and you continue to pay for monthly subscription, the more money they make, and that is really the core of the problem. This distribution problem is completely solvable if you have the intent to solve it. But when your own business KPIs actually conflict with what your own users want, that's how we end up in this dramatic environment.

Speaker 1: 31:35

Basically what you're saying. Let me just translate that to make it clear. Say, I'm in the bottom 10% of men on these platforms, and again, we're just talking statistically. We're not trying to suggest that the bottom 10% are not worthy or anything, but if you're in the bottom 10% of men, then it's very unlikely, statistically speaking, that the top 10% of women will want to match with you. Right, that's the reality.

Speaker 1: 31:58

So if I were a responsible app, I would reduce the likelihood of you seeing these profiles, and if you wanted to like spotlight with them, then I would warn you and say hey, look, you know what.

Speaker 1: 32:10

This person is probably not so likely to want to match with you. Just keep your roles and probably match it with someone else or something along those lines. Right, I'd be more responsible about this stuff, but obviously that's not within their business model, ok. So, with all that said, though, I would like to also point out not only are these apps statistically a bit lopsided and weird and obviously create these weird dynamics, but now I think there are also behaviors that have shifted in society and say I am a woman that's getting inundated with likes, comparatively, because there's 75 of these people men on these apps that are liking me well, I feel quite self-confident to an extent, and then when I do match with a guy, right, how, what is my attitude right? And maybe you can tell me, as a straight woman perspective, how you've been seeing this or talking to women about this if you start from the woman experience.

Speaker 3: 33:03

I use dating apps in my 30s, both in london and new york, as I was living between those two cities, and what would happen is I show up in Manhattan, I open Bumble and I would get 2000 likes under 15 minutes and you know for those of you who don't know me, trust me I'm not in the top 10% of women in Bumble. The reality is that's the female experience, and then I would feel overwhelmed. What am I going to do now? What I want to filter by is is this person into technology? Because I'm a complete nerd and I really enjoy talking about technology. Has this person been well-read?

Speaker 3: 33:45

It's quite important to me to be able to have interesting conversations over dinner. I don't have filters for any of that. What filters do I have? Their height, their, you know, how many children do they want? Those factors, while maybe very important for some people, have zero importance in my selection of my partner. So I would start actually checking a couple of profiles. I would probably get through 1% of all of those, but even that would be still like 15 people who are writing back to me. It's a Thursday night, I'm tired, I got back from work. If I actually had a partner who shares my interest. Of course I would love to sit down, have a glass of wine and have a good chat, but do I want to meet a stranger who just wrote me a couple of lines? I really don't know anything about them. No, and even the best of us end up behaving in a way that we wouldn't do in physical, normal life.

Speaker 2: 34:44

Tell me about that, because that's what I've heard from most women who've come off the apps.

Speaker 3: 34:50

Completely like on my side. I would, for example, not respond to people for 10-15 days, sometimes like never, and that is what's called ghosting. Your intention is not ghosting, you're just too busy or too overwhelmed to go through. You are unable to choose. It's a decision paralysis. At that point I'll give you another experience.

Speaker 3: 35:16

A good friend of mine was going to go out on a hinge date on sunday and he was quite excited about it. We were out somewhere else and then he called me later on oh, can I join you guys? I I was like what happened with the date? And apparently she just messaged him literally 30 minutes before the date saying that look, she was a software developer in a very large tech company really interesting person, so I'm sure she's a good person, right. And she said I'm feeling terrible about this, but I really don't feel like going out. I'm feeling pretty dark. You know, I really don't feel like going out. I'm feeling pretty dark, you know. Please, like, accept my apologies, maybe another time. I don't have the power. Why is this happening? Because you've been A on so many dates that were disappointing. Because there is no pre-date preparation in terms of are we a good fit? Do we have some common interests. What are we going to talk about? And that disappointment accumulates over time. Picking yourself back up and showing up yet another date becomes really hard work.

Speaker 1: 36:25

Yeah, and on the male flip side is you just get disappointed spending so much time on these apps because women will cancel on you or they go on one date with you and then they have 500 more likes. So between the first date and the second date you have to contend with 500 likes per day that they get and you're like have to stand out and keep attention and like get people to come, come back. And if there was one thing they didn't like in that conversation on that one date, then they're like, okay, there's somebody better that doesn't say that among these 3,000 people that just matched with me. So let me just go for that. So there's always the better person around the corner. The abundance of choice.

Speaker 1: 37:02

And disposable affinity yes, disposable affinity is the right word here, yeah.

Speaker 3: 37:06

That was my exactly second point that this is the right word. Yeah, that was my exactly second point. That's an illusion of choice because the actual options are not there. 30% of profiles you are seeing on a mainstream dating app have not been active for a year. Unless you actively go to an app and deactivate your account or ask to be forgotten, they continue to show your profile to people because they want to create liquidity, right? So the reality is you get this illusion of oh my God, I have so much choice, but actually the translation of that is not oh my God, lots of abundance, I'm so happy. No, no, no, no. It's more like I'm so overwhelmed, I'm paralyzed, I don't know with whom to move forward.

Speaker 1: 37:55

Yeah, exactly Okay. So we've established the minefield that is current dating apps. Let's talk a little bit about and I think you were alluding to this earlier the fact that you know we should be coached a lot more here and we should be trained a lot more here. There are people like, for instance, matchmakers, and you know, you've spoken to some, I've spoken to some, and it's really interesting that they're trying to create a different side channel now, in this world where there's abundance of profiles that are thrown at you, to find someone for you and talk to them directly with a different operating system than these apps basically create. Right, they're like games that these apps create, and this is a way to step out of that game and play a different game matchmaker is an interesting concept.

Speaker 3: 38:40

They did exist way before dating apps did. It goes to very earlier jew communities.

Speaker 2: 38:49

It was very common in Jewish communities.

Speaker 3: 38:52

It definitely was very common in India. Later on we've also seen it, probably potentially an export of India to the US and UK. We started seeing it in the US and UK. What most people think when they hear the term matchmaker is a probably tall blonde woman based in Mayfair. You know have a book of people that they can get through and they basically introduce people that are handpicked for you. Typically these cost around $30,000 minimum per year. It's an incredibly expensive service that is only available for very privileged rich people very much.

Speaker 3: 39:33

So the interesting part though you know you can afford it, you go buy the service, no problem with that. What I find quite interesting is more than one person mentioned to me, as a female, how they found the whole experience of working with matchmaker even offensive, because what matchmaker did in those cases and these are very well-known Mayfair-based matchmakers trying to manage their expectations that they should lower their expectations. If you're a 35-year-old woman in London and you still want kids, then you should pretty much agree with anyone that we are finding for you, and I mean these women are high power, fit, take care of themselves really well. Trust me, they have a lot of options, and hearing that from someone that you pay $30,000 a year to is incredibly unacceptable. So I think it depends. Like anything, where it's human run, there's a huge variety of experiences.

Speaker 1: 40:40

Absolutely so. We're basically saying there is an epidemic on the dating app side and there's people that are trying to kind of fill that gap. Matchmakers are one of them. I've also seen a rise of all these relationship coaches I mean plenty that are now even posting on linkedin or other platforms, which are professional platforms, about how you figure out what you are searching for really and how you can then find the right partner, etc. Tell us a little bit about that. I'm sure you have looked into this absolutely.

Speaker 3: 41:09

I think there's maybe two camps on that. We have esther perels and qualified people on the one side. Every time I listen to her, it's a, it's a pleasure and on the other side we also see the instagram tiktok versions of that from people who has absolutely zero qualification, not from a personal experience nor from an academic degree that we should be listening to them. But they are articulate and they commercialize it and they use the right trending music on TikTok and it works right and all of a sudden you are getting advice from someone just because he got 1 million likes. It is becoming a very big space.

Speaker 3: 41:45

Now the reality is let's put aside the. You put aside the charlatan side of the industry. The real side of the industry is actually serving a very big need, because being stuck in the wrong relationship can take away any sort of joy and happiness from life like nothing else. Changing jobs are easier. If you don't like your flat, you get a new one right. It's to some extent easier, but actually being stuck in a wrong relationship is very hard. So teaching you, understanding yourself what your needs are, coming into terms with communicating your needs and understanding the other person on the other side, so that you can get closer to each other, for an existing relationship is incredibly important. Now let's walk back from that. If we know that, why don't we do some of that pre-work in advance? Let's use the technology to work out what are those compatibility factors that actually make people together happier, and why don't we lighten the burden that you currently have with having to go through thousands of profiles and instead only show you the ones with which you have a good chance of being compatible?

Speaker 1: 43:08

Perfect. I think this is a great segue into.

Speaker 3: 43:11

Alchemy and Luna and AI.

Speaker 1: 43:13

So how do you think AI can help us with this?

Speaker 3: 43:17

I think in a lot of ways, ai has been helping for a long time. So, for example, being able to sift out fake profiles right Again, we talked about 30% inactive, there's another 30% which is catfishing and fake profiles right Again, we talked about 30% inactive, there's another 30% which is catfishing and fake profiles and being able to identify if someone's a genuine person or not with AI, is becoming actually quite easy to do. What generative AI brought in is something a little bit different Because with Gen AI personalizing not only a machine learning algorithm but an entire language model and a bot hotel to go to you know how do we do our flat search, but definitely in one of the most important areas, how we choose our partner.

Speaker 1: 44:16

Okay, you had a good point around flaws.

Speaker 2: 44:19

Yes, it is interesting that the moment there is a bump in the road, someone just wants to get out One of the things for me in a relationship I saw my husband, as you know, in a 360, a very holistic way, I think, and he did me and I or he I can only speak. What he did for me really was he sort of amplified my, my daily life and my experience of um any kind of interaction, and not just giving me a sort of binocular rather than a monocular experience of life, because his perspective was always very different than mine. He wasn't irritated by the things that I found deeply irritating about myself. So, whatever my fault lines are, whatever my character defects and my flaws were, he forgave and vice versa, the things that he loathed about himself I just found quite endearing and I found it funny that he found them so offensive. So there's so much elasticity, I suppose, and forgiveness on those sides that from what I hear and from what you guys have been saying, and even in conversations when friends are trying to set people up this sort of what do you want?

Speaker 2: 45:32

Is the first question what do you like? And give me your list of unacceptable and give me your list of aspiration and what you really and I've just the one thing that's all sort of missed is surely, if you're connecting on those things, you're not dealing with any of the sort of subterranean stuff that comes out much later. And how does? Can an AI possibly access those flaws and then find what people say is an absolute no-go for them? Or where you suffer from low self-esteem, let's say that's not necessarily something you're going to put on an app as part of your character, but if you admitted it to your AI bot assist, maybe they would. How would they?

Speaker 3: 46:15

And that's what I'm curious about is I'm skeptical about there's so much to unpack in what you are saying, but you so beautifully described the experience you had with your husband. That really brings us back to what does top 10% mean? There is no such thing as top 10% If you pay a photographer and you edit with AI some of your images and you can be in that top 10%. But the reality is you have your own personal top 1% and sounds like you found it Because for that person, what you consider to be a flaw for yourself, they are not necessarily even forgiving it. They just don't see it as a flaw. I had the same experience for many, many years.

Speaker 3: 47:02

I dated so many people who are hardcore avoidant and for those of you who are not familiar with the term, we talk about attachment theory as a theory in psychology that our childhood bonds can make us either a secure person in future. It can make us very anxious about the connections we have, questioning how strong that connection is, or it can make us avoidant that with the fear of rejection, we basically rather avoid getting into relationships. Right, and I'm an anxious person and the thing is there's a trauma bond between the anxious and avoidant people. You actually attract each other, but then a lot of things I don't like about myself in the first place becomes exaggerated in a mirror and it effectively makes the relationship not really work in the long term. In a different type of relationship, where I feel safe, being with someone who is very secure, very comfortable in his own skin even the things that I thought was really problematic about me I got to rethink. Reconsider that Even your love for yourself, self-compassion, how you feel about your job, what you can achieve in the world, can be significantly enhanced. That's why we are, at Luna, so passionate about what we are doing, because if you get the right person in your life, that experience cannot be substituted by anything. Now, how are we doing this with AI? Right? That's really the question, and the first thing to say is it's a learning journey. No one can claim today that we've got all the data, we've done all the analysis. This is how it works. There is no such thing. We are co-creating Luna with our community. We are learning from them. So, as they are sharing their experience, we did something very unique that no dating app did, but because they couldn't do which is a feedback loop they couldn't do, which is a feedback loop.

Speaker 3: 49:09

When we introduce you to people. The first thing is you start by telling us quite a lot about yourself. We ask you about things that caused your previous breakups. It's a little bit like a therapy conversation with your dating coach, luna, and it's not about asking you do you have low self-esteem? I mean, let's be real, most people may not be even aware of that, right, let alone admit it. But instead things you say to us how did you make your travel choices? How much you were influenced by other people versus made your own decisions? How much risk you took in your life in different stages of life.

Speaker 3: 49:44

We ask you all those questions and then there's a psychological framework at the back which calculates your scores on self-esteem, anxiety, being an avoidant, how independent you are and you want to be, how compromising you are. Now, this is the beautiful thing. There is no such thing as being higher or lower on the rank. It's a fit right. So if you're a very secure person, we are going to match you with people who would really appreciate your security so much that some other things that may be problematic about you is not going to be on the top of their agenda. They're just not going to care right? That's how we define. We turn every user into a vector. That's a multi-dimensional universe vector. As a vector, you carry numbers around your independence, your compromising nature, your anxiety and like hundreds of other factors like that and we start finding the closest vectors to you.

Speaker 3: 50:49

We introduce you right and then we ask you, did we get it right? You talk to that person, then you give us natural language feedback, not just yes or no. You tell us that, look, I really like that person, but he checked his watch twice while we were talking. That made me really anxious. I don't think he's into me Now. We got so much from those three lines.

Speaker 3: 51:14

Not only now Luna can step in to give some coaching to you, because clearly you are feeling anxious and you can use a little bit of a therapeutic conversation. That someone checking their watch could be due to a lot of reasons. Maybe they were really stressed about their work deadline or something right. So Luna is able to deliver you that coaching to make you feel better in your own skin. But she also taken a lot of notes that you need to be with someone who is emotionally attentive so that they would realize you feel anxious when they check their watch and would match you with someone next time, who has very high emotional intelligence, so that he can cater better to your needs.

Speaker 3: 52:00

Obviously, we are biased. We believe in the power of generative AI and what it can do, but to me, the most important thing is, no matter who you are, there is a right person for you. If you got into a relationship with that person, your life is going to be uplifted. Our job is, with your help, with all your feedback, to sift through all the you know in the haystack. Sift through all the you know in the haystack, sift through all the others and then bring up to you those 10, 15, 20, whatever it is that we think you have a very good chance of having a successful relationship with.

Speaker 1: 52:35

Yeah, so I was lucky enough to be part of the beta test and this was the early product. Right, it's not in the final stage, but to give a perspective from somebody who has used it, who is not the person running the company, you answer about 20 questions about yourself and what it does. Then it creates a profile for you and there's only one image attached to that profile as of now, and then it basically shows you some profiles and tells you what did you like about these profiles? What didn't you like about these profiles? So then you read about other people's profiles that are also, by the way, generated by ai, but are much more comprehensive than anything you would ever find on any other app. Right, there's like really good information about somebody, and then you give your feedback.

Speaker 1: 53:21

So this feedback loop is something that does not exist on any freaking app right now. None of these apps has any understanding of how my, my, my matches went or not, and the fact that it doesn't have this feedback loop means that it's always day one. It will never do anything differently from the first time. I opened this app for me, and that's silly, right. So even my definition of that it's it's great to know. Okay, this app is actually listening to me, it's giving me some feedback. Now, all of a sudden, it gave me different matches, and then you know so it's progressive it's progressive exactly, and that's, I think, the beauty of it.

Speaker 1: 53:58

I already feel it, it's absolutely unique, yeah, I mean, as far as I'm concerned, like all the old paradigm apps anyway, there may be some new paradigm apps that pop up, but like all the old paradigm apps, they're all silly. And you see the ceo of bumble now saying, hey, we need to get into artificial intelligence as well and we can talk a little bit about them. But ultimately, I think this clearly seems like the future and there is a moment here for us right now.

Speaker 3: 54:21

Do you want to maybe talk a little bit about bumble and what I just said, I think your day one example is a very, very good articulation of exactly the problem. Right, I want something that actually learns about me and works for me as my agent to get me off this platform as quickly as possible, because that's really what I want and that's exactly what we're trying to do Now. I think we're at a stage now, do you remember like for those of you who are old enough 20 years ago, we were seeing some businesses becoming like digital businesses, or there was a department called digital marketing. It's the same thing happening right now. No business will survive without incorporating significant amounts of generative AI in the next two to three years. So if Bumble will go ahead and include some of this technology in its current app, it must. Otherwise, it's not going to survive. Now, when you look at the Bumble's makeup, there are no people in their team who are LLM specialists, nor there are any open roles even in that direction.

Speaker 1: 55:30

And the stats that we mentioned still apply 75% men, and you know they still apply.

Speaker 3: 55:35

But I think that there's actually two more important structural problems, because with goodwill and right leadership you can make things happen right. But I think that is the problem, because what are you going to do about your business?

Speaker 1: 55:48

models right.

Speaker 3: 55:49

The moment you're actually successful in matching people, they will come off the platform and your current business model just does not align with that, and that's quite a big change. There is a second big structural problem. To be able to match people, you need to know about them. I learned so much about both of you today that I didn't know before and now if I was actually trying to match you with any of my friends. Now I have a much better understanding, like what you're looking for. So the conversation needs to move away from hometown. How many children do you want? What's your height?

Speaker 1: 56:27

Vaccinated or not? Vaccinated or not? Star sign.

Speaker 3: 56:31

Oh my God, or are you a beach bum or like a triathlon runner? I'm not saying these are not completely unimportant, but they do not actually determine the success of a long-term relationship. So what is Bumble planning to do to learn the level of your self-esteem and how confident you are in presenting yourself? What approach are they going to have to make you feel more confident in a date? That's, to me, the more interesting thing to see.

Speaker 1: 57:02

Mm-hmm. Yeah, I think the point that you're saying you have a hundred dimensional model of a human being while they have a five dimensional month is kind of the core problem here, right absolutely.

Speaker 3: 57:13

And the interface. We are living in a whole new world. Lunad runs on whatsapp.

Speaker 2: 57:19

Yeah, we don't have an app right, we are basically natively gen ai solution I know there's tons of personality tests or go back to sort of iq tests, kids aptitude tests at school, and the meaninglessness of an awful lot of those things and how unhelpful um, they've been in categorizing people as certain things, that then it's taken them a whole life to kind of shed the shackles of that.

Speaker 2: 57:45

So I'm not a massive believer in these structures that are fairly absolute and believe in themselves. I'm more interested in what I think is maybe possible with generative AI is the flexibility it will allow rather than the yes, no answers, as you pointed out. But I wonder how capable it is of nuance or of reading someone's protest about a certain part of themselves. As we say, I never told a lie, I didn't believe myself right, so I'm liable to lie about myself in order to put up a positive front. How do we hack through that? Because, actually, in getting to know someone's fault lines and finding them inoffensive and acceptable, that you can then build all the other bridges and I think there are two points there you're making.

Speaker 3: 58:39

One is I completely agree with you that it worked. We were all categorized into you're an ENTJ and you are this and that, and you know my profile was an ENTJ. I'm a hardcore introvert.

Speaker 1: 58:52

This is Myers-Briggs, by the way.

Speaker 3: 58:54

Yes, it's Myers-Briggs, and you know I used to be a consultant and we always had to get our profiles. The way I act in business may come across as extrovert, but actually, like, I get my energy from, you know, much smaller group conversations, etc. That's why at Luna, we are not doing profile quizzes and we don't give you a profile type. It would be so easy for us to do that, but we are completely, you know, moving away from that because humanity is so complex. Not only you will be so different to Omed in 100 different dimensions. Even your dimensions will be different.

Speaker 3: 59:31

It's not just where you are on that score. What defines you, what matters to you in a relationship, is going to be significantly different. And that's the beauty of the compute power. With GPUs we have, we can make that level of compute at an affordable cost and at a very fast speed, and we can store that information. The other part is that's again why we come back to we did start using some psychological theories. For sure, we used some frameworks to improve our models at this early stage, when we don't have sufficient data yet. But the rest comes from the people, and I love the beauty of how people reveal themselves. I think that's a beautiful thing, Like when you realize that when you are in a relationship with someone, if you've been together for a year, all the curtains at some point go down right, Like you actually really get to know them. Why? Because we do need to be ourselves in the end of the day.

Speaker 2: 1:00:27

But you know, we have as many personalities as we have friends, so we change depending on who we're with, and how you might behave in one relationship is entirely different from how you might behave in another, because someone is pulling out of you a part of you that they are particularly linked into, and so that part of you grows, and in another relationship that part of you may have wilted and died. You can have a relationship which can enhance certain parts of your personality and you can have a relationship which maybe even sows seeds for other branches to grow in you that you didn't even know existed. Hence going back to the thing of you like yourself more when you're with that person, because it's growthful that, rather than getting in the way of your growth, they're actually creating growth.

Speaker 1: 1:01:15

But you don't know those things until you're in relation to that person, I think to make it simple, I know exactly where you're going with this, and the reality is that human interaction is probably a thousand dimensional space. Right, that's what you're saying. And it's not only a thousand dimensional space, it's also always changing, and it's changing in a chaotic way, based on the causal cones you find yourself in Other people influencing you and work and like hormones and the you know work circumstance and all that.

Speaker 1: 1:01:48

Now there are parts of you that are probably much more continuous and other parts that may change. I think that's probably a true statement. That said, I don't think tucci is suggesting and you may correct me that this paradigm is perfect to begin with, but it certainly has the potential to become more perfect than anything that currently exists, right.

Speaker 3: 1:02:12

It's the fact that Luna, or the model, evolves with you. That's the most important part. And the difference is if you look at eHarmony or OkCupid, they also learned about your personality. They asked you a whole bunch of questions. The problem was exactly what you are raising. It was static and it was based on self-declaration. Even for people like us who studied maybe psychology etc. There's so much you don't know about yourself. Or I come out of a relationship with my guns up because my independence was so challenged. Now I'm saying I want someone totally independent, I can't be clingy. And then I meet someone and I'm like okay, like the clinging with this person doesn't feel so bad, so I change my mind. Right, that's what the feedback loop is for. We use something called parameter, efficient fine tuning and adapters Say that again.

Speaker 3: 1:03:10

Parameter efficient fine tuning and adapters which allow us, cost efficiently, to keep changing your personal model based on what you just said, in the last two weeks, four weeks and we tend to not change it just in the last week, because you might be just having a bad day, you might be just having a bad week, but eight weeks in a row if you keep coming back after your dates or after your matches and telling us that, oh look, I actually feel quite nice around someone who is very protective and wants to do a lot of things with me. I thought I really liked independent people, but you know what James last week was fantastic. I want to see more James type of profiles. We're learning from all those experiences and maybe you go into a beautiful relationship for three to four years and it's, you know, runs this course and you come back Again. You may be this time looking for something completely different. That's the beauty of generative AI models that it evolves with you.

Speaker 2: 1:04:13

Because you're quite evangelical about this, or it felt like it was almost a sort of social cause. Is your devotion to this because you think it's superior to meeting someone in real life?

Speaker 3: 1:04:24

I think my devotion comes from the fact that we need to fight modern loneliness, and I think that's becoming an epidemic. If you go ahead and meet someone in real life, go for it. You are very lucky and you should enjoy that. But for the rest of us, who are the majority, who is a very big part of the population let's use the technology we have to offer you some options to live the rest of your life with someone who makes you feel great about yourself, who enhances your life, and I think dating apps changed the humanity for the worse. It brought the worst out of us.

Speaker 2: 1:05:07

Is that because of the anonymity element?

Speaker 3: 1:05:09

Anonymity is very important. Illusion of volume and too many people are a problem, being overwhelmed, losing self-esteem because of constant rejection they're all part of the problem. Where we are now, we had enough and you can see it. Even the declining user numbers these big dating apps have. We are seeing actually a mass exodus now and that's basically allowing us to be able to enter at a time that we can give people a very different dating experience.

Speaker 2: 1:05:45

So you're meeting a demand.

Speaker 3: 1:05:48

And that's the reason, probably, with a soft launch on a LinkedIn post, we just had so much love and interest coming through. Most people who are joining us are saying that they are done using dating apps. They don't expect to meet anyone who meet their criteria on a dating app and we are now giving them a completely new approach where Luna trains and learns how to be your coach, how to represent you and how to choose for you, so she does the heavy lifting and then you can actually show up to dates where you already know so much about the person. We always say in the office we start people from the fifth date. Right, we can sort out all the stuff that you would have gone through for the first four dates. So you just go enjoy a glass of wine or whatever it is that you like and talk about deeper things. That gets you closer with another human.

Speaker 2: 1:06:44

What's your price model?

Speaker 3: 1:06:45

I'm just in terms of you talked earlier on about democratizing this and how, with those matchmakers, there's these crazy amounts of money exchanging hands in as feels transactional almost sort of dishonest, actually not transactional that they're taking your money and not really giving you much in return yeah, we believe strongly that a very big percentage of our user base is going to be free, so we expect people to be able to come and use Luna and enjoy it, and that's very important because even for the paying users, it's important to have a large pool of people that they can find the right person within, and we are very happy for people to come and use the solution completely in a free way.

Speaker 3: 1:07:31

We will have paid users, we will have premium users, and that's primarily for people who are poor on time and they don't want to actually go through the profiles. They want Luna to do all the work for them arrange the dates, they just show up. For those it will be $900 one-off fee, so Luna does not profit from you staying single. You pay Luna only once as your dating coach, she takes it over, she schedules the dates for you, you just show up, and for people who are really time poor, that's an option they would want to use I like um the pricing model because one of my heroes, who unfortunately passed away recently, charlie munger, always said show me the incentives and I show you the outcomes.

Speaker 1: 1:08:22

And when the incentive is, you know what? If somebody stays on the platform and costs me a lot of money, then I want to get them off the platform as soon as possible, instead of the other way with the current dating apps, where as long as they stay on it, they continue to spend and therefore I'm having a much higher lifetime value of users who stick around for longer and even for those who are paying $900, it might come across as quite an investment, but what we always say is how much are you planning to spend on your wedding, don't you want to make sure you are marrying the right one?

Speaker 3: 1:08:57

In my view, that's a worth investment to make sure that you are merging your life with someone who is going to give you lifetime happiness.

Speaker 2: 1:09:06

But in the case of your generative AI, the reason why you will have free subscription is because the more data you're getting from people and the more feedback, the better and the more accurate the AI becomes right. It's self-perfecting A hundred percent.

Speaker 3: 1:09:23

I think that's where we keep talking about co-creating the platform. Without our community sharing data honestly with us, we don't have a business. It's as simple as that. For us to be able to work for them, they share what they want, what they are looking for, how they feel, and then we use all of that data transparently, with their full consent, to train models, which not only learns to find them better matches, but also learns overall what makes people compatible in general.

Speaker 1: 1:09:59

That's a great point, I think, to end this interview on. It was really exciting to look at the future of dating. I am going to be using Luna for sure, and I'm looking forward to revisiting this at some point in the future to see how you're doing.

Speaker 3: 1:10:14

Amazing and right now we are not accepting users, but you can always go to myalchemyai and join the waitlist.

Speaker 1: 1:10:21

Awesome. Thanks so much, petr. Thanks Tüce Thank you.