Description

Journey into the Future of Human Behaviour Analytics with Hayk Hakobyan. If you enjoy deep dives into the intersection of technology, human behavior, and business innovation, this conversation with InsightGenie's founder will captivate you.

‍

In this fascinating episode, we sit down with Hayk Hakobyan, Founder and CEO of InsightGenie (formerly BizBaz), to explore the revolutionary world of behavioral analytics and predictive human assessment. From his unique background in nuclear physics to pioneering voice-based personality analysis, Hayk shares how his company is transforming the way organizations understand and predict human behavior. Learn how advanced AI and behavioral science are reshaping everything from banking and insurance to HR and recruitment, with accuracy rates exceeding 90%.

‍

Transcript

‍

Jonathan Nguyen: 0:24

Welcome back to the Unsensible Podcast. We're on the road today. As you can see, we're not in our usual studio. It's very dark in here. I think it's just meant for audio, but I'm here today with Hayk Hakobyan. Hayk is the founder and CEO of InsightGenie. We're going to talk a little bit about it. The topic is going to be quite wide-ranging today. Hayk. What is InsightGenie? We're going to talk a little bit about it. The topic is going to be quite wide-ranging today. Hayk, what is InsightGenie and what do you do there?

‍

Hayk Hakobyan: 0:51

It's actually funny. You say what is InsightGenie? Because until probably six months ago or a year ago we didn't really have an InsightGenie. It was called BizBaz. It still is because a lot of people know us.

‍

Hayk Hakobyan: 1:02

But what we essentially do is we offer I'm going to go with your lines or maybe not yet no, we offer basically customer intelligence and eventually, as Jonathan here would say, we offer confidence to make the right decisions and right choices.

‍

Hayk Hakobyan: 1:21

Whether you are a bank trying to onboard a customer and deciding if that customer is the right customer for you or not, whether you are a bank trying to onboard a customer and deciding if that customer is the right customer for you or not, whether you're an insurance company offering an insurance policy, so we offer the kind of data and intelligence that will help you decide with confidence and with reliability if that's the right customer.

‍

Hayk Hakobyan: 1:38

So we enable and empower those power, those organizations to onboard and work with until then, or either to unacquirable or unbanked or uninsured or et cetera type of customers. And also recently started working in the HR kind of domain where pretty much the same thing Again, when you're recruiting a talent, we want to offer basically the right type of data and the confidence that comes with the data that, yes, this is the right talent and you're not going to be wasting your team's performance, et cetera, when you recruit a bad talent, right, which is very common nowadays. So, yeah, that's kind of what we do. So we offer data and, let's say, the confidence sorry credits to you confidence that comes with it to make the right decisions between organizations and individuals.

‍

Jonathan Nguyen: 2:29

Tell me a little bit more about your background. I was looking at what you studied and you're literally like a nuclear physicist.

‍

Hayk Hakobyan: 2:41

I studied physics, I studied business, I mean academically, and then I picked up some other stuff along the way, some of the human sciences as well. I'm just, let's call it, I'm a kind of human being. I wish a lot of us would be just curious, which is what we were, let's say, 100,000 years ago. Right, we were curious and wanted to learn, and that's what it is. I'm trying to bring back that early curiosity that humans had.

‍

Jonathan Nguyen: 3:05

That's what it is. I'm trying to bring back that early curiosity that humans had, so that curiosity has taken you on quite a journey right. So you started out doing your PhDs in nuclear physics, which is fundamentally like maths degree. How does that happen?

‍

Hayk Hakobyan: 3:20

I think the underlying and a common thing here is that I was just generally very curious as a person I've always been before I started studying it. I just like to understand stuff. That's actually even the reason to consider physics as a line of study, because I just wanted to know stuff. That has to do a little bit with my kind of upbringing and my grandmother was a teacher and still didn't ask these children the curiosity. So I wanted to learn stuff. I thought physics could be good. I didn't realize it's very not applicable in a way like you don't get a high paid job for this. And then once that was done and I had the degree and then I thought well, um, I started getting interested in human sciences simply because I realized that physics, as much as it is well known, well documented, etc. But it's not really able to explain day-to-day or not even day-to-day. Just look around, I mean, how society works. Physics cannot explain that. There are some mathematical models that are now basically trying to explain, like complexity theories et cetera. But even beyond that, if you want to go into micro level, understand how individuals work and you deal every day with individuals right, it's a society you want to get ahead in your career, in your life, you want to get things more efficient. You have to understand how people are, not on a group, societal level, but individual. You're like, yeah, it's, but then I got it. I'm like, yeah, but this, all this great science doesn't explain how I can talk to this person. How do I get him to do what I want him to do, or how do we efficiently communicate and get something together? So I picked up and I picked up the if you think of all this natural science, human sciences, as an iceberg, I picked the tip of it, which was gamification at the time. So I picked it up and I like games. So I'm like games. How to deal with humans. Okay, that sounds good.

‍

Hayk Hakobyan: 5:00

I learned it and I was in Mozambique at the time. So I was taking the class when I was in Mozambique and I was the VP of innovation and strategy in what was, at the time, the largest opposition media in the country. So, on this, it's like a media company, right, they have print, they have digital and we had distributors and I was like taking a class and we had distributors. These people only distribute once a week and we don't really get any reporting done because it's a big country and it's very poor, but we would love to. Then I thought, okay, how do I practice? How can I use gamification to gamify the distributors, like, how can I get these people to do what I want? So for me it was not theoretical at all, not academic, not nuclear science, but basically, okay, I know I learned something. I want to use it in my daily and that's how kind of it started.

‍

Hayk Hakobyan: 5:49

Right, I didn't read a lot of books and then theorize and write a lot of LinkedIn posts. I just I think I was like two months in a class and I went to our founder I want to make sure my goal is to get news from ground up, because that's what the position media was about. He said, great, how are you going to do that? We've been trying to do this for years. I said how did you do that for years? He said we tried everything. We tried to even offer them more salary. We tried to incentivize them, give them salary, give them some material stuff.

‍

Hayk Hakobyan: 6:15

I said okay, and he said, no, it didn't work. We have 50 distributors, not a lot. How about? How about I have a go at this? I'm taking this new class. Maybe there's something there. He's like go ahead. So he led me Guinea pig, basically the distribution team, and I did, and it was very successful. And that was literally I haven't even finished the class and I already had a success and I got like, wow, okay, this works. And I don't even know anything about it, I'm still kind of learning and still works. And so we gamified the distributors. They started basically reporting news and then, because of its great success, we also kind of put it out there into the digital like because the newspaper was very kind of well-read et cetera, and I haven't stayed long enough to check the records, but it seemed like it was working On kind of a back of that kind of initial success. I started kind of going more and more and reading about it and just practicing and advising. That's how it.

‍

Jonathan Nguyen: 7:11

And then you went into a new Went and studied neuroeconomics, yeah, so among others.

‍

Hayk Hakobyan: 7:15

I studied neuroeconomics and a little bit after. That actually was quite interesting because I had a friend who was running an orphanage in southern Tanzania which is bordering Mozambique friend who was running an orphanage in southern Tanzania which is bordering Mozambique. So I was very excited and I said, hey, how is it going? And I was going to catch up with her and she said, yeah, but I have a hard time. And without me even saying anything kind of coincidence, she said I have a hard time.

‍

Hayk Hakobyan: 7:33

All this, like I forgot those 80, 90 or 100 kids in orphanage and she was trying to get them to learn better and also trying to get them to also kind of do the course and stuff right. And she shared some of the challenges. I said, how about? I'm going to come because I was visiting Tanzania. And I said, how about, like you and I, we meet and I give you some ideas.

‍

Hayk Hakobyan: 7:52

So I went and again, this was a couple of months after I finished the class I was in Tanzania for climbing Kilimanjaro and I visited her in some remote village in Southern Tanzania and I said so I took this class and I'm doing this and I see some early success. So for kids, here's a couple of ideas you can do to gamify the course and the learnings and stuff on that. And then I think, like within six months, she said oh, I'm like I don't even know anything. I just took your advice, fleshed out what that's going to look like, the bullet points you gave me, and I'm really happy. I mean, I can see the results. Maybe I should go read more about this myself. I'm like great, I'm happy.

‍

Jonathan Nguyen: 8:26

You apply a lot of that now.

‍

Hayk Hakobyan: 8:30

I apply a lot of that now in my family setting as well as professionally.

‍

Jonathan Nguyen: 8:35

Tell me a bit about how it works and how you approach. When you use the tool you speak into computer yeah and it gives you a bunch of results. Now, if someone hasn't ever tried it before, you know, I tried it at a conference can you tell us, yeah, yeah.

‍

Hayk Hakobyan: 8:56

So I mean, we are essentially in the business of predicting human behavior, right? I mean, as I mean, I think that's the part where that magic thing comes in, because you don't think you can. This is not a mathematical algorithm or some sort of a computer where you can predict what's the next step, or even a chessboard, right? So our business is business of predicting how a person is going to react or do or repay, or behave or perform, right? That's where we are One of the components where we get the data from about the person is voice, like you perform, right? That's where we are One of the components where we get the data from about the person is voice, like you said, right? So it's different components, but one of them is voice. The reason it's becoming quite known is because it's quite cool, right? Quite magical, as you say, because people don't think that their voice has anything about them. So, voice your laryngo, basically the part, the apparatus that kind of generates the voice. Interestingly enough, the science says that basically after your puberty whatever that would be 18, 19, 20, it starts concealing specific characteristics and even things such as, for example, if you have depression or stress, et cetera, that gets somehow concealed in your apparatus, in your physiology and an untrained or even trained person, like in this kind of conversation, you would never know it. But if you look at, but if you evaluate using what's called the prosodic analysis of your voice so basically just physiological or biomarker analysis of your voice and if you have an algorithm that is trained to do so, based on the science that exists, then you can extract things like, for example, or calculate things like how conscientious the person is, how impulsive this person is, how emotionally stable that person is or isn't, and even obviously psychometric things like how open he or she is, how collaborative, et cetera. In a way it's like a Pandora's box, right. If you know how to evaluate your voice from a biomarker perspective, you can have a very good kind of a window into the personality of that person or the character of this person.

‍

Hayk Hakobyan: 10:49

And imagine suddenly, I mean you know a lot about this person, I mean you don't necessarily know what will be the next step, right, you know he's conscientious as this, he's impulsive, et cetera. You are able to in some ways gauge potential behavior, future behavior, like if you see for, for example, a person who comes very high on, for example, impulsivity, and he's not emotionally stable and he seems to already have mental kind of problem. You don't necessarily know what mental problem is, but there is something there. Then again, you don't have to be a behavioral psychologist, behavioral scientist, psychologist or even a therapist to know that if I'm gonna to offer this person life insurance policy, then in some ways it becomes almost like intuitive once you know these things how it is. It's not, but like intuitively. You know that if someone is high on, for example, impulsivity, the kind of reactions, the kind of behavior he or she might exhibit, right Versus someone who is very stable, low on impulsivity, and if you're a bank, then you know, right, okay, is this person impulsive, but he's relatively emotionally stable, okay, so impulse buying is a thing.

‍

Hayk Hakobyan: 11:54

How about I offer this person a credit card that would fit his character or her character? But I also make sure that interest rates are. And if the person is, for example, conscientious, I have a peace of mind and that's why I'm talking about the confidence business, of confidence, right, like you said, I know that this person, however impulsive he, she, is, will pay and in time. But I know something this person is impulsive, then obviously you can always talk about the philosophical, moral side of things, about this kind of thing. But point is we empower one side to serve and engage with the other side organizations and people. We provide insights and a lot of them seem magical because we're trying to bring order out of chaos.

‍

Jonathan Nguyen: 12:36

The voice is one data set.

‍

Hayk Hakobyan: 12:38

Yeah.

‍

Jonathan Nguyen: 12:39

But you would probably in that situation, mix a number of different data sets together. And maybe it becomes a probabilistic type of it is yeah. What is the extent of what you can do and what is the extent of the datasets? You mentioned the banks. They could take a voice dataset. What other type of use cases are you seeing?

‍

Hayk Hakobyan: 13:03

Yeah. So when we started, there were other innovators, who are still there and it's great that they exist and they do what they do. We decided to take, I feel like, a more holistic approach to this. Versus being specific to data, which some of the others are. We thought from behavioral perspective, it doesn't really matter, as long as I can get some data that's actionable and based on the science, I can figure out what that data should be. Then we should be good to go.

‍

Hayk Hakobyan: 13:29

Now you're right that as much as it is not very deterministic I mean so it's not deterministic what we do obviously it's still probabilistic because I predict, with whatever level of confidence, that you are going to be, for example, repaying your loan or your insurance or you're going to perform really well in that organization from HR perspective, it's still probabilistic because it hasn't arrived yet. It's futuristic, right, it's in the future and, unlike the mathematized, let's say, sciences, which even then I mean if you play chess it's still probabilistic, right, even though it's very deterministic, the steps there is no uncertainty on. You do this and some stuff will happen that you don't expect. There's only so much. It's just. It's always. I think I would think of it. It's always probabilistic, even in chess, because there is only a probability of, let's say, 20 steps, 20 things that can happen in five steps. It's the same thing here I'm predicting future and as it hasn't arrived, by definition it's probabilistic.

‍

Hayk Hakobyan: 14:30

We try to make the whole behavior or character of a human as not determined that would be a bit weird, right, but as accurate as possible. And we can do this. For example, for voice, the accuracy of our predictions is, on average, 80%. That might not seem like 100%, but that's already super high. And then some of the other things. I mean if we had, for example, voice component and we try also to understand this person's mobile phone related behavior using mobile phone number. Still not 100%, but in theory 100% is not achievable or attainable, right, because human nature, et cetera. But we are coming as close as probably you could with some sort of technology. So that's where we are, I think.

‍

Hayk Hakobyan: 15:08

So we try to, as you say, use different data sets and understand human character, understand human behavior, and we do it contextually. This is super important and the reason I mention is because you see all this like big five for Oceania. Whatever personality types, myers-briggs, et cetera A lot of them have one assumption. In most cases and I'm not saying this lightly in most cases, context is more important than your own, so to speak, characteristic. It's not just a factor or it's not within the context. No, no Context dictates how you will behave. Our assessments are all contextual. So context evaluation is super important because that's also going to dictate how you behave. And again, I mean, we are still at the earliest kind of stage of this. We haven't been around 20 years, just a couple of years. It seems like this is the right way of assessing right.

‍

Hayk Hakobyan: 15:55

You look at dispersed or disparate attributes of a person using voice, using mobile, using online internet, and it's scary for a lot of people. And the funny thing here is that a lot of people, for example, they don't know how burned out they are or what's their conscientiousness. It's the first time for them to discover that about themselves. I'm not even talking about like, yes, you know what's your financial history, the bank will discover it, but you know exactly what you have.

‍

Hayk Hakobyan: 16:19

This I'm talking about here where, like whoa, my conscientiousness is like three out of 10, really. And of course, people who are at that level they might say, no, I'm not. I'm actually super conscientious, which is kind of confirming right, people are conscientious. I'm not going to argue in this kind of way, right or like my impulsivity is nine out of 10. No, I'm not impulsive. They just exhibit exactly what the thing yeah, so it's kind of, and obviously for the bank is the same thing. But the challenge also and I'll just preempt you on this the challenge with this kind of data is it's relatively new and not well known and hence most of the organizations are not equipped I would say probably almost all organizations don't have the right data scientist type of skill sets to deal with this data.

‍

Jonathan Nguyen: 17:12

It's historic reasons, but the idea is that, yeah, if I give you a person's heart stroke probability or I give you some most conscientious level, how that connects into their risk model is a very complex problem. Yeah, and there's very. You know, I recently did um life insurance, filled out a life insurance form uh, it wasn't life insurance, just health insurance but the the questions they ask what kind of workout do you do? What do you eat? They can just say do you smoke, do you drink, and it works, but it's a very, very crude measure. Then you must come across lots of these types in what you do now. I mean, we spoke about credit scores, for instance.

‍

Hayk Hakobyan: 17:44

The reason why credit scores are insurance-related risk assessment the way it's done. It's just historic because, lines on history, first credit score so to speak, technological was invented by FICO so the two founders of FICO in the US in 58, and credit scoring or risk assessment generally as a technology became mainstream at the end of 70s, beginning of 80s. So that's when banks until then it was like this an interview I would sit with you, I'm the credit officer and I'd say, Jonathan, tell me what do you want? Yeah, do you want a loan? Show me some collateral. Normally it would be collateralized. Show me you have gold or you have car, you have something, and then he or she would just subjectively assess if you are worth that. And then he or she would just subjectively assess if you're worth that.

‍

Hayk Hakobyan: 18:29

So that became technology in 1980s and in 1980s I'm talking about beginning of 80s what do you have about yourself that I can assess? If I'm a bank or insurance or anything? The only information you have about yourself that is assessable is your banking history or your insurance history. So that was just your financial history in 80s. Right, you don't have internet, uber, facebook, none.

‍

Hayk Hakobyan: 18:51

So that's what it was used, because that's the only thing there was about yourself, because the bank can provide financial history of a few and then you go to the next bank and that became the bureaus, right, and then fast forward 40 years and it's the same thing. That's because and we don't have to go into that because they're conservative, a lot of points Now you have all this richness of data, the internet, uber, facebook, voice assessment, video assessment, et cetera. I would venture I guess 90 plus percent of all financial organizations still don't use it globally, due to the nature of the financial industry. How it is. Should they use it Definitely, is that future? Yes, how soon you are going to be on that train to the future depends on how innovative you are, or?

‍

Hayk Hakobyan: 19:37

how much you need it. Or how much you need it If you're a large bank, without naming some here, and you don't care about, let's say, working with unbanked or underbanked, you probably don't need it. But seeing as it is that most of, let's say, asian countries are by and large, unbanked and underbanked, you probably want to start tapping your feet into that a little bit. Or every other bank is looking at it. The question then becomes okay, do I and I give a lot this laptop analogy, but the question becomes for organizations do we build our own, so to speak, risk assessment profiling inside customer, inside system, or do we use a vendor? Unfortunately, unfortunately, most organizations think they can do it themselves because they think this is a tiny, important but tiny piece of their overall machinery. It's like, again, you take an Apple laptop, you're like so should I produce my own CPU or should I kind of get a?

‍

Jonathan Nguyen: 20:31

vendor.

‍

Hayk Hakobyan: 20:31

I think I can produce one, why not? And they spend a lot of time and effort and money and resource. Most of them fail and they fall back on the existing one, the credit bureaus. Great, they exist, credit bureaus, but again, a lot of the reasons they are still in business is because, due to the nature of financial industry not being very innovative and being super slow and a lot of data-related limitations things are moving at a pretty slow pace in most places and that's the reason bureaus still exist in their form, in their current historic form.

‍

Jonathan Nguyen: 21:02

Do you see so traditionally when you see technologies that get adopted later? So let's say, in developing countries, yeah, and now I would say that you know, most homes probably wouldn't have a desktop, they'd have a laptop. Yeah, if you go into more developing countries. Maybe they don't even have that, Maybe they've gone straight to mobile.

‍

Hayk Hakobyan: 21:26

Yeah countries.

‍

Jonathan Nguyen: 21:26

maybe they don't even have that. Maybe they've gone straight to mobile. Maybe they've skipped the whole fiber and landline and DSL and gone straight to mobile technologies? Do you think that at some point they won't even use the traditional credit score system or the weighting of the traditional credit score will be a lot smaller in these countries that are going to go behavioral first?

‍

Hayk Hakobyan: 21:51

Yeah, that's a good question. I don't necessarily have a good answer, but because credit scores, however underdeveloped the countries are, I think one of the first things in pretty much anything that's a country, even things like East Timor or South Sudan right, the newest countries you have banks. The moment you have banks, you would have a bureau. One exception, though Bangladesh. So Bangladesh is the only country in the region and it's a very big country, about 70 million people. They don't have a bureau. They have a credit database, which is a precursor to a bureau.

‍

Hayk Hakobyan: 22:28

I don't know so much about African countries, to be frank with you. Sub-saharan Africa, all of the North African countries do have bureaus, and it also doesn't help that all the big three right Experian, equifax and TransUnion they've been pretty aggressively expanding, and the smaller ones as well, globally. So to find a country without a bureau or a bureau data or like a database kind of precursor to bureau, is going to be very hard. I don't. I would venture a guess it probably doesn't exist. If you have a bank, there's a good chance you have something that provides financial history. It's almost like one of the first things as a country or a city that you have to have.

‍

Jonathan Nguyen: 23:04

What about the usefulness of that if, like 80 or 90% of the population does have a bank account?

‍

Hayk Hakobyan: 23:10

Well, that's the thing, right. So if you look at so, what happens? The current state of things is, if you look in Asian countries, they go with alternatives. So that's why, like for example, in Philippines, wallets like GCash or coins, etc. They're super popular. Or OVO, et cetera, in Indonesia, because Indonesia is 270 million people and 80% are unbanked and 50% of that don't even have smartphones. So you have wallets. So the gap is filled by substitutes. Of course, substitute doesn't mean everyone else is going to be qualified to that, but big chunk of population that missed on that banking thing is using that Remittance wallet payments and stuff. You go to villages in Indonesia or Philippines or whatever, and you see people have the probably cheap, low-end kind of smartphone and they have wallet one, wallet two and they have a little cash here, a little cash there. So the alternative, so to speak, financial system is there, unfortunately mostly built by wallets that are now very popular because of what they do. They cater to an unmet demand.

‍

Hayk Hakobyan: 24:17

Bureaus are now waking up to the fact that a big slice of their cake is being eaten because wallets the trends in Asia are like. Wallets are now looking at becoming more of financial institutions. It's both sides right. Wallets try to become more like banks. I mean, take even Grab. Right, they have started with a wallet and then they added insurance, debit, credit cards, and that's the trend.

‍

Hayk Hakobyan: 24:42

You have a wallet, it's basic, it's functional, a lot of people use it. But then people who use it, they want more stuff, not just the wallet. How about an insurance, a card, I don't know mortgage, et cetera? And then banks on their side, they're like well, everyone uses wallets, even those who don't have a bank account. How about we enter that segment by offering a wallet? So they go top down okay, here's a wallet and, by the way, we are a bank, we have all this other stuff as well. So in that way again, innovation happens from ground up and it's mostly by like, let's say, the wallets or payment systems. And if you want to revamp this and start like a new, here is an alternative credit bureau, which what we were called at one point by TransUnion, where it's all behavioral data, or behavioral and financial data, et cetera. That my guess is that won't be, because you cannot scrap, like you say okay, let's scrap all this, start from scratch. That won't work like that, unfortunately. It has to be gradual, incremental, and status quo exists already in every country.

‍

Jonathan Nguyen: 25:43

I had the opportunity of speaking to your head of tech and he described the interview process and I can relate.

‍

Jonathan Nguyen: 25:52

Like you know, before I was startup, I was in a very big company and those interviews you're interviewing a lot of people- yeah you're asking a lot of questions and you ask it over and, over and over again and then you finally get this person on board after three to six months and then you start to discover things that didn't come out in the interview. So he described his pain. Tell us a little bit more about how you see, you know that kind of behavioral model, mathematical model, of being applied to the recruitment process or the HR process.

‍

Hayk Hakobyan: 26:32

So we have what you recently dubbed as behavioral quants, right? So this is basically. These are data guys who are obviously versed in using financial data, but also behavioral data, to crunch what would be traits of personality or traits of character. So this whole agr thing started kind of quite organically. I mean, we had a couple of banks were working with and I remember I got the question once and then we started looking internally how to like apply this and found that some of the technology we have could actually be applied. But question initially was could we like it's a banker, can we also recruit team members using this? Because what you give us essentially is behavioral characteristics of a person. Sure, we can onboard new customers, but I want also new team members, right, and can we use that for that?

‍

Hayk Hakobyan: 27:24

Initially we were a little skeptical because we thought, well, I mean. And then we realized something. I mean it's the same building blocks. I mean when you're recruiting someone, like we do, like they do and everyone does, beyond the technical. I mean if you think of, let's say, level of expertise, you're looking in people and, let's say, a career ladder, right. If you look at very junior, the most important thing is their technical ability, right or their aptitude or whatever ability to pick that up. As you go high up the level, let's say all the way to C level, their technical knowledge is becoming almost not required, right? What is required as you go up the ladder in your recruitment is soft skills, team play, mental well-being, conscientiousness the stuff that constitutes a big chunk of what we provide anyway for customers to onboard, for credit card bureau whatever, et cetera.

‍

Hayk Hakobyan: 28:19

So we thought, okay, so we're no HR experts, but having our own experiences, like our CTO told you, and then looking around a little bit, you don't even have to dig very deep, Like, just look around, I mean, everyone is. And again, like I said for credit bureaus, that we're giving a historic thing. It's exactly the same thing. The fact that you have this kind of interview is historic because there was no other way of doing it. Like, what are you going to do? You want to recruit someone in 60s or even all the way to 2000,. What would you do? You sit with someone, you ask some questions and you use your subjective judgments to figure out if the person is good. Now you have some tools, for example video-based assessment of your big five personality traits.

‍

Hayk Hakobyan: 28:58

But I commented on that right? I don't think it's very accurate. And it gets perpetrated because there's no negative feedback. You don't get fired after three months of being recruited using Big Five because, well, lo and behold, you are not as open or conscientious as they thought. That's never happening. You get fired because you don't perform but no one really does the connection or the feedback to whoever there was that initial so appealing, simple, logical, completely misleading system in some ways, yes, it will.

‍

Hayk Hakobyan: 29:28

If you're a big gambler, there's a good chance. I read somewhere that Big Five has, I think, 30% 40% accuracy in predicting what you are. The usual way is the status quo, way you talk, and then sometimes you have to recruit more HR people to recruit, like, for example, in construction industry and stuff like that, and assessment is very subjective, Unless you're a very, very experienced HR person, in which case 80%, 90% accuracy is your prediction. But how many of those are there? They are costly, they are rare. Everyone else, yes, they can have an impressive resume, a couple of large companies but unless you have 20, 30 years of HR experience, your predictive accuracy is probably not that good.

‍

Jonathan Nguyen: 30:08

That's the other thing as well. That's interesting, though and I'm tying this back to something you just said is that if you've got 20 or 30 years experience, you're not interviewing. People are you?

‍

Hayk Hakobyan: 30:17

Yeah, probably you're not right. You're probably a big headhunter, one of the big top headhunting companies, or you're part of Unilever. You're the HR director. You're part of Unilever, you're the HR director, but everyone else is. And there is all the subjective assessment problems, unconscious biases about people that they don't even know. Like, even if you're a behavioralist or behavioral scientist, you're still afflicted with the same stuff, right? Even if you know that you have an unconscious bias, you are still subject to it because it's your design, right? You can't help it.

‍

Jonathan Nguyen: 30:47

Sometimes it's training as well. Sometimes they teach biased training.

‍

Hayk Hakobyan: 30:52

Yeah, exactly.

‍

Jonathan Nguyen: 30:56

Someone walks in for an interview, you're supposed to look at what they're wearing and their shoes and you know like it's not even a training.

‍

Hayk Hakobyan: 30:59

It's how we are evolutionarily wired to take all the signals and on the subconscious level, we already like. Yeah, it's not a training. There is a lot of research that shows, for example, and you probably know this, but for the audience, better looking people get ahead in their career from get-go, from junior to like, better looking, whatever that means White people, white get better results, get higher Men, right. There is all kinds of unconscious biases for genders, for race, for how you look. How you look is, to me, the most shocking. The better looking you are, the further you go. I mean, the kind of correlations that exist is just mind-blowing to me. And the best part is none of this is deliberately done. That's because, again, evolutionarily, we would choose humans just like the rest of the animal world. We were built to initially take signals and even if you don't know what beautiful is or it's subjective, you know what you see, right, and a lot of people could agree what it is. So you choose, based on some signals, what would be the ideal mate for you, or ideal partner, ideal this, and that we are still subject to that made for you, or ideal partner, ideal this, and that we are still subject to that and we don't want to admit it, but it's the same.

‍

Hayk Hakobyan: 32:11

So the moment someone walks in, how they look, how they sit, how they're dressed 70%, 80% of the kids it's already done. They sit, they open their mouth and it's 100% done. It doesn't matter what you say. The rest is you go through the motion. Of course, from that moment you can, if you say something really bad, you probably go two or three steps back. But if you come back and if you come in and you already I don't know you look very, you don't look very attractive, you're not dressed as what's expected from you by that person. You start here, it's an anchor, so you get anchored the moment you walk in and then from that anchor it's this range or that, yeah. So what we don't know, what they don't tell, is that you're already anchored, you don't even have to open your mouth. You walk in and out and they have an average of you. Then they're like oh okay, fine, fair enough. But it's all about amplifying and just making the whole process more reliable and faster, because time takes money. Time costs money.

‍

Hayk Hakobyan: 33:06

So one final question for you, and this is a hard one Inside Genie does amazing.

‍

Jonathan Nguyen: 33:12

And 10 years from now, what does that look like? What does success look like? What's the impact on the world and business?

‍

Hayk Hakobyan: 33:21

I think in some ways success if we did well. There's a lot of uncertainty and who knows right. Odds are always against us, so it's much lower probability we'll do so well. But if, let's say, we defy the odds, it would be that you would have basically one company, or ideally, because if it is successful, a lot of the others would enter that field and try to replicate or whatever.

‍

Hayk Hakobyan: 33:44

A lot better way of assessing human character. It's both good and bad. Good because if you're in a society and you want to deal with each other, et cetera, it's going to be a bit like this dystopian fantasies. So people, you cannot even tell them there is no freedom of choice or free will and stuff like that which is very high, very abstract. Imagine if you tell someone well, I know you're this, this, this, this, this. If I give you a loan, there's an 80% chance you're not going to pay me back. Telling people about this kind of stuff it's kind of undermining the way they think of themselves, because telling themselves you don't have a freedom of choice, for example, and stuff which we can argue in another time, but that's already pretty like threatening to their personality and identity. And, of course, if we do well doesn't mean you're going to walk around and people are going to come to you and say, no, you're this, you're that, no, no, but people will know that bank knows a lot about them, or insurance, or the company wants to recruit them, or I don't know marketplaces, and in some ways I mean it's bad because it could threaten, but it also, in the other way, it can incite people to be better. Perhaps in some ways, if I take second, third order thinking, if I know that I'll qualify for this and this is also psychology right. That's why you have leaderboards. Leaderboard is not a threatening thing if you know how to do it Points, badges and leaderboard right, the PBL system. This is what we did.

‍

Hayk Hakobyan: 35:22

By the way, going back to Mozambique at the beginning, when we started, I would put a list of the distributors. Every week, one day before distribution. I'll say here is how you did, and I would literally sit back and see everyone come and look at that list and say, oh, you sent 10 messages with news. I sent five. How come? Like? Where'd you go? Like? It incites competition. As a result, the overall system benefits. So it could happen like that, but that initial moment it was not pleasant and it's probably going to be the same here. It won't be pleasant, but if it's done at scale, on a societal, whatever level, if again we do well, hypothetically, etc.

‍

Hayk Hakobyan: 36:01

I would say human character would overall probably start improving, because people here don't have any drive to generally don't have a drive to improve. Why would they? I can get a job and I'm happy and I get my salary. But if you know that your salary will depend on your character, because character will dictate your performance, if you know that you'd probably try to improve because you would want that indirect and kind of same. And then, of course, if you do leaderboards and there'll be more visual to help drive that. But the idea is that, yeah, if you know that your character is going to determine the quality of your future, quality of your future you would want to do something about it. That's an incentive, that's an intrinsic incentive almost.

‍

Jonathan Nguyen: 36:47

There's also something there about also understanding a much broader. You mentioned context, right.

‍

Hayk Hakobyan: 36:54

Yeah.

‍

Jonathan Nguyen: 36:55

There is something there about. You know, if you want to be an event manager, right and you're trying to be a banker, but your personality is more suited to event manager yeah, you're going to fail at this job. Yeah, correct this job.

‍

Hayk Hakobyan: 37:15

Yeah, it's going to be better for you and so it will be that kind of course correction, that because right now, when you're going into, let's say, university studies, you don't really know what you want to be. I mean, in the best cases you're saying, oh, my dad and mom say this, or this is where the money's. You invent some criteria or you tell yourself some stories like that, whether you'll materialize as one of those. But if you had this, you would probably someone say, look the way you are right now, you're 21. The building blocks like concession stuff. It indicates that you'd be more successful, more fulfilled and happier as a consequence.

‍

Hayk Hakobyan: 37:50

If you're this, I would want that. Yeah, I would want someone to tell me, or system ideally, say go here because, yeah, you go, you go, try and fly. Right, you cannot, don't have the physical kind of a machinery for that. You'll fail, right. I was saying that says if you try to judge how high a fish is going to fly, it's going to be a failure, because fish is not built to fly. Most fishes are not right. So you tell people, here is where you should play along, and sure, that doesn't mean you cannot go here and there is no way for you to learn. Sure, but it's a longer way and it's steeper to learn, sure, but it's a longer way and it's a steeper. So at least you know.

‍

Hayk Hakobyan: 38:31

Yeah, everyone will be better off, right, because not everyone is going to be becoming a lawyer. Some people go a lot and they go being a meditation teacher and they are super happy. They said this is my calling. Yeah, wouldn't it be better if you just go straight there? I mean, it could be. You can always say argue that the process blah, blah, blah. That's how they are, and but it's always a good to at least know in advance a little bit. You can still go to the law and earn a lot of money. Then go there, but it won't be a random drunkard's walk, as they call, right, yeah, so awesome, great hike, thanks very much thanks a lot for insightful, insightful questions and that's our that, and that's my phone that I forgot to turn off.

‍

Jonathan Nguyen: 39:09

It's a good signal for us to leave it there. So thank you very much, thank you everyone, and we shall see you in the next one. Thank you.

‍