My Pomp Podcast Interview about Relationships, Rationality, Startups and Bitcoin —Video and Transcript

Pomp was kind enough to invite me on his popular podcast last week. He did a great job asking questions about the kinds of stuff I write about on this blog.

Click here for an interactive video+transcript combo. I’m also including the transcript here.

Pomp: All right guys, bang, bang. I’ve got Liron here. Thank you so much for doing this, sir.

Liron: Hey, my pleasure. I’m a longtime listener.

Absolutely. Well, we’ve known each other for a while now because I had the pleasure of investing in the company you’re running right now but we’ll get to that in a second. Let’s jump right into your background. You’ve got a pretty interesting story where you’re born in Israel and moved to Silicon Valley at the age of four. Just talk me through growing up in Silicon Valley and what that was like?

Sure, yeah. My first language is Hebrew but I got a chance to go to kindergarten and stuff in the US. Pretty quickly, I got Americanized and me and my brother started talking mostly English. I was just really lucky to land in Silicon Valley because I love computers. Then, later I discovered I love entrepreneurship. I looked around and like all the cool groups and organizations were just right here and 30 minutes’ drive away everywhere I wanted to go. Yeah, I was like obsessed with computer programming growing up.

I studied Computer Science in college. Then, right out of college, I worked at Slide, which was like, at the time, I had just a random job in San Francisco, Slide but then in retrospect, it’s like, “Oh, wow, there were a ton of really great people there who were my coworkers, obviously, Max Levchin, Keith Rabois, Suhail from Mixpanel, Dora from YC, just like crazy amount of good talent density at Slide, and I didn’t fully appreciate it at the time.

Then, I went and did a startup called Quixey, which itself got really big and partnered with Alibaba but actually, we ended up kind of lighting $170 million on fire that basically had a terrible outcome. That was kind of my foray into the startup world. Then, a few years ago, kind of picked up the bat for another swing and that led me to my current company and Relationship Hero.

Let’s talk about lighting $170 million on fire. Explain kind of what happened there. I think for those that aren’t familiar with startups, the idea that you could be a fast growing big company and still not work out is a little counterintuitive. Kind of elaborate on that.

Yeah, totally. That’s a really important idea. It’s really interesting to think about. With Quixey, we got really big and this was, we started all the way back in 2009 when the iPhone was new and the whole idea of app stores renew. Our thesis was like we need to invent a totally different search technology that’s customized for finding the right app.

Now, I actually think that that’s thesis, there’s still something to it that’s under exploited. If you use your iPhone, if you pull down the search bar or the spotlight feature, it’s actually still really crappy. It only recently started finding the most obvious results and it’s still like barely, if you type in like the name of a document that you’re working on, it’s not going to pull up the document. That’s still like a rich area for improvement. That was kind of the seed of Quixey but we ended up kind of flailing around trying to find what’s the revenue model? How do we get traction.

Our biggest breakthrough came when we partnered with Alibaba. We made this deal where we would power search for their app stores. Then, that turned into like a 20 million a year revenue thing. We were doing a bunch of custom work for them but eventually it all fell through.

Throughout the course of that whole process, we raised mostly from Alibaba as a strategic investment, a total over seven years of $170 million. You think if you’re looking at it from the outside like, “Wow, this company is going great and all this money’s pouring in,” but a lot of times, it’s an illusion and you really just have to look at the yearly metrics like how much revenue is coming in? What are the costs? Ours were pretty crappy, but we just had all this money to burn anyway. We just burned it and then we shut down because we didn’t even have an asset that was good enough to sell and continue as independently.

Yeah. What’s so interesting about hearing this story is the way that you run Relationship Hero is very, very data-driven. I get the updates and it’s probably one of the more data driven updates that I get. Talk through a little bit about where did the idea for Relationship Hero come from and then we talked about the park you guys have built?

Totally, yeah. You’re absolutely right. We have we do AB test for data driven. The original idea, it really has nothing to do with my previous career but it has to do with a problem that I understand deeply, which is struggling with dating and relationships because you always want to start a company around a problem you understand. As somebody who’s always been very technical, living in Silicon Valley, like my career has always been an area of my life that I have done pretty well but my relationships were like a disaster. I’d go on a date and it’d be like the worst day. I feel really awkward. Communication was bad. I wouldn’t get a second date. I find myself always hitting up my friend like monopolizing his time, getting all this advice about dating.

After my last company shut down, I was thinking, what problems am I familiar with? I’m like, “Well, what if we take my friend skill, which is basically relationship coaching, and we scale it up, and we give other people the same benefit that I’m getting,” which sounds a little crazy, but as we started to explore it, it’s like, well, everybody’s trying to talk to their friends. They’re trying to get relationship advice from their friends. It’s not like they would even bother contacting a professional because professionals out there just have like these random websites, they’re not really trustworthy. They’re just different types of coaching. There’s a lot of therapy, but therapy is or marriage counseling but that’s not exactly what people need either.

We just thought, hey, maybe there’s a niche. That was like the initial kernel of the idea. We kind of grew it organically. The first version of Relationship Hero was five of my friends sending us screenshots of text from their Tinder, from there, okay, Cupid just being like, “Hey, what do I do? What’s my next move?”

As you start to build the company, it sounds like really, in the beginning, it was almost like we’re going to do the service and then we’ll figure out what the product is over time, how do you go from what is a really early, early kind of MVP to what the product has become today? Is that just constantly talking to the customers and really understanding kind of how they’re using whatever product they have in their hands and what they need or is this more of, you’ve got a vision, you set that vision, you go build it, and then you’ll find the customers that want that thing?

It’s a combination. It’s a concept that I call “following the value” or I call it the “lean MVP flowchart”. You’re always trying to follow a flowchart in terms of what’s your next step. The rule of thumb is like, well, where can I double my value? How much value am I creating now and just how can I double it because if you keep doubling it, pretty soon, you’re going to create a lot of value.

For example, like at the beginning, we were helping a few friends with TextMasters. I was like, “Okay, well, to double that, I just have to find five more friends.” Pretty soon, I’m helping 10 friends. Okay, but then I ran out of friends. Suddenly, I had to get creative. I’m like, “Okay, well, how am I going to find 10 more people?” We went on Reddit and we found some people on Reddit, and we’re helping them for free.

I’m always just thinking how do I just double and now we have thousands, so doubling thousands. Now, we have to think like, “Okay, well, now we need to do a TV commercial or something,” but if you just keep asking, how do I double, how do I incrementally create more value, it gets you pretty far.

Yeah. As you start to actually build the product, I remember even early on, you’re playing with the monetization and kind of the pricing structure and that will put it was like, I think $1 per minute and then you move more to like a monthly-type structure, just talk to you maybe some of the tests that you would run and kind of what data you would pay attention to in order to better understand what works, what doesn’t work and kind of what’s in the best interest of our customers long term?

Totally. Yeah, one thing we do is we just we AB test like crazy. Anytime you launch anything, we have it deeply embedded into our software that everything is an AB test. I’m not just talking about like button colors. A lot of times we actually don’t bother testing like the really small stuff that we don’t think is going to matter but anytime we’re trying to say like, “Hey, can we raise our prices?” Well, what if we were to offer, we raise our prices, but we give half the people a discount and we see if the people with the discount actually tend to sign up with a higher rate.

We’re always looking at the numbers because how lame is it to like launch something and that cuts 25% out of your business, and you don’t even realize it, and you blame it on seasonality. You know what I’m saying? You need the instrument panel of AB test.

Yeah. This idea is pretty straightforward, right? It’s, hey, people want a relationship advice, we’re going to provide relationship advice but you’ve built it in a unique, almost like marketplace way where you’re bringing in top of funnel interest of people who need the advice, and then you’re basically finding people who could be the coaches and onboarding them in order to provide a good experience. Why has nobody else done it in this way? Where did the insight come from to do it in this specific kind of virtual technology enabled way come from?

Yeah, definitely. Relationship Hero, it’s not necessarily what it seems on the surface. When people see us, they’re just like, “Oh, you’re a marketplace,” but what most people don’t realize is that there’s not really a market for relationship coaches. You can compare us to BetterHelp and Talkspace on the therapy side, those apps are much more like a marketplace because they take something that’s a more standard commodity. Like a therapist, okay, you make a list of what certifications it constitutes a therapist, and then you go offer a contract job to those people. That would be more of a marketplace model.

In our case, there’s no such thing as a certified relationship coach with a single definition. We looked at all the certifications out there and we’re like, “Okay, well, this one kind of fades, but it’s missing this,” and eventually, we pull together our own certification and our own training. Then, we’re like, “Okay, great. We know how to train and certify relationship coach, but then where do we find them or like who are these relationship coaches?”

We started kind of looking for raw talent. We have people. We have a list of like 20 backgrounds that are all somewhat relevant, like, “Okay, this person has done social work. This person has a BA in psychology.” We have all these different backgrounds and then we train them. We’re kind of in this middle space. I guess there’s a fancy term for this for like a managed marketplace where we have a lot of control over our supply. In fact, in the US, the coaches are actually full time W2 employees.

Got it and so as you started to scale this up, what’s been the biggest learning in terms of actually building out, in a space that people I think understood, but using a model that people hadn’t done before?

Yeah, it’s really just thinking from first principles, right? Instead of, I feel like people are always like trying to search Google and try to find the answer for how to do something, but in our case, we’re just like, “Well, can this person give good coaching and if they did, how can we tell? What would be the proof?” We made a bunch of tests. They have to take tests and then we have a dashboard showing what is their retention look like? Did the clients come back? What kind of reviews did they leave?

When we hire and train a new coach, we can see very quickly, like within a couple of weeks where they stand and we can instantly detect anomalies being like, “Okay, something’s wrong with our coaching.” The unique thing about what we do is we take responsibility for the quality of the coaching. If you got bad advice, we’ll actually go in and welcome back, “Okay, let me give you a refund. Let me actually give you a more senior coach to kind of fix the bad advice you got.” It’s rare, but we take responsibility when that happens.

Why is the focus so much on what I’ll call like, very tactical things, right? This isn’t, “Hey, be nice,” or “Hey, you should go do something at a high level.” It’s very, very tactical and you should respond in this way or you should do X thing on Y date. Why the focus on the tactical aspect rather than more high level stuff?

Sure, yeah. Well, imagine you’re going to your friends for relationship help, a lot of what you need is that tactical advice. The thing is that it’s possible to be an expert on relationship tactics. Courtship is a game, right? It has a lot of evolved strategies and tactics. Like crafting the perfect text, you have to draw on a lot of communication skills. Everybody feels like they have to be their own expert, right? Everybody’s like, “Oh, yeah, I don’t need help. Dating and just comes from my heart. It’s just me being authentic.”

One of the perspectives that we have our Relationship Hero, which we think is not super popular yet, but we think it’s growing in popularity, is the idea of like, “Look, sure you want to be authentic, but you shouldn’t expect yourself to have all these communication and courtship skills because it’s something that you can invest more time into getting better and you can talk to a professional and get some tips that will save you a lot of time.”

The most important question I’m going to ask you today, everyone wants to know, how do you write the perfect text message or Tinder message?

Yeah, definitely. I’m not going to give you a bunch of stock openers, although we do have those. The number one key that people just don’t get, the most common complaint I hear about dating apps is like you send a message even you work hard on it, and you express why you like the other person and then you don’t get a reply, or you got a few replies, but then the conversation like Peters out, and you’re like, “Damn, I like this person, like the conversation Peter out.”

The number one trick to keeping the conversation going in a really good serendipitous way is you just have to think a couple moves ahead. Think about it like it’s chess, like playing chess against yourself. Imagine that you are the recipient of the text and imagine that you just want to be able to respond within five seconds. A response just comes to your head like it’s natural banter. Most of the texts that people write are actually really annoying to respond to and the only way to respond to it is to think for five minutes. When you make the other person think for five minutes, they’re like, “This guy seems pretty cool but let me get back to him,” and then they never get back to you.

Yeah, it’s so fascinating that you describe it as chess because basically, what you’re trying to do is not ask a question that has a finite answer, it sounds like. Instead, what you’re doing is really asking a question to elicit a response in which you can then respond to again and kind of keep the conversation going, right? It’s almost like you’re optimizing for a sustained conversation rather than simply getting information.

Yes, except when you say sustained conversation, it’s not just about prolonging it, it’s about making each back and forth feel good basically or give some value or have some purpose. For example, when it’s early in the conversation, a lot of times you just want to make them engage and have fun. Your goal is basically to create a fun interaction. You want to throw in like a teaser or a joke, but not just like a random knock-knock joke, but like a certain type of joke where they like it and also they’re motivated to continue it, something that challenges them a little bit, kind of like an interactive role play.

I love just the framework that you use to describe this. I want to switch gears and we’ll come back to relationship here in a little bit but there’s a lot of people who listen to this who like Bitcoin and technology investing. You’ve been early to a lot of things. There’s this tweet from 2011, June 11, 2011 and it says, “My investment thesis: Buy into a 10% probability of 100x return. Examples one, Bitcoin to Milner and Conway’s Start Fund.” There weren’t very many people who were talking about Bitcoin at 100x potential upside back in 2011. When did you first hear about Bitcoin? Why did you kind of tweet that out? What was the logic back then as best you remember it?

Yeah, definitely. By the way, don’t forget Milner and Conway Start Fund, remember Yuri Milner and Ron Conway back in 2011 gave Y Combinator a ton of money. They were invested a lot in every company. That ship has actually come into, right? If you look at all these IPOs, like Airbnb, almost 100 billion. Everybody now realizes that Y Combinator is by far the best game in town except maybe Sequoia. Those two are… Yeah, I mean, I think both Y Combinator and Bitcoin were really sleeper hits back in 2011.

Where did I get the thesis? I just think that the human brain is just bad with orders of magnitude, right? I’m always looking for the 10% probability. Back in 2011, Bitcoin, there was like a 90% chance that it was just going to fizzle out, right? Now, obviously, it’s much more robust but in 2011, sure, it may fizzle out but I just noticed, “Okay, well, on the 10% chance that it doesn’t fizzle out, couldn’t it just go 10x, 100x, 1000x? Why isn’t that possible? If it can go 3x why not 100x? Sure enough, the 100x was like well within the realm of possibility.

Yeah. In terms of that technology versus a fund, how did you look at the difference or even today, how do you look at the difference between basically giving money to somebody else to go invest versus finding the assets yourself?

Sure. Yeah. I mean, they’re both good strategies, right? I tend to, I just think that my comparative advantage is noticing, I call it order of magnitude perspective. My brain is just wired in such a way where I’m always trying to ask how big is something? I’m just like really obsessed with order of magnitude size. I just noticed, “Hey, Bitcoin could be really big, right?” You know that very well. You’re always talking how it could be 100 trillion one day. Yeah, that was the deal with Bitcoin.

Also, one connection I have to Bitcoin that’s kind of random, what was I even doing looking at it in 2011, a couple random historical accidents from my life, one is that when I was in college studying Computer Science, I randomly took this grad class, graduate level elective and we were talking about cryptography. I randomly did a project, writing a little paper on this research paper where somebody had made a scheme for it was called Compact E-Cash. It was like a way to use principles of cryptography to pay people.

I did a little report on the scheme. I did a little presentation. I’m like, “Yeah, this is a nice little scheme but it’s impractical. It basically sucks, but it’s like a nice idea.” Then, literally like two years later, I’m reading the Bitcoin white paper and I’m like, “Wait a minute, this is like the crappy paper that I read in college, except much better solving all these problems.” I just randomly happen to be in a position of being like, “Oh, holy crap. This is like a solution inborn problem.

Then, the other thing that really hit for me is I was part of the lesswrong.com community, all these rationalists and they were actually very early to Bitcoin. In fact, there’s some gossip that one of the contributors, one or two of the contributors to LessWrong might be Satoshi. They were talking about Bitcoin. The last straw for me after those two data points was just seeing the market price start to rise, like seeing it get traction, seeing it get talked about and not die. I’m like, “Okay, if on the conceptual level Bitcoin is so powerful and it’s not going away socially, it’s actually picking up steam in terms of like normal people caring about it, then that is like a very potent combination.”

You also were an early investor in Coinbase. For somebody who saw Bitcoin so early was it just a natural extension when you met Brian that Coinbase could be big or kind of what was the story there and why did you invest?

Randomly in 2012, when I was kind of like at the height of my Bitcoin obsession, I’m actually less involved with Bitcoin these days but right around that time, there was this new company called FundersClub, which is really great. They’re actually doing really well, too. It was this new idea of letting people give smaller checks, a little bit like a syndicate, letting people put smaller checks into companies. That was right when I was starting to get a little bit of disposable income and I was looking to do a little angel investing.

Just by coincidence, I ended up putting a check into the first FundersClub fund where they had a few options of companies to pick. I’m like, “Oh, yeah, Coinbase. I like Bitcoin.” A lot of luck. Silicon Valley being at the right place at the right time happened to Bitcoin, ended up angel investing in Coinbase and then, of course, forgot about it for a few years and now they’re like the next potential YC IPO. I’ve just been really impressed. My thesis there was just like, “Look, I think Bitcoin is a big deal and I think that Coinbase is doing a really good job just being like the best Bitcoin company.”

It’s so funny because once you make the investment, you kind of forget about it for a while. It’s almost as if you give your capital to an entrepreneur and you test them with like, “Hey, go grow my wealth,” to some degree. Being in the YC Silicon Valley type world, was there additional weight put on it that they were going through YC or had gone through YC, I don’t know at what point or was it just, just like this company?

The YC stamp of approval definitely meant a lot to me. YC has a special place in my heart because I was reading Paul Graham essays when I was in high school just like obsessively, even before YC was a thing and then I kind of followed YC from afar, read all the articles documenting what was happening in YC. Eventually, Relationship Hero ended up doing YC in 2017. I kind of got to check off a bucket list item but yeah, when I saw Coinbase do YC, that was definitely another plus.

But at the same time, it was really funny to consider how clueless I was as an angel investor because Coinbase is my best angel investment. I like to act like I know what I’m doing but it was really my first lucky new investment. I was just like chatting with them being like, “Oh, yeah, let me analyze your code. What part of the [inaudible 00:20:44]? Just didn’t really know how to evaluate a company.

It’s so funny. What was going through YC like?

It was pretty cool. It’s just I built up so much anticipation. I read so much about YC that when it actually happened, it was slightly anticlimactic because it’s like, “Oh, yeah, right. Okay, it’s YC.” I felt like I was super hardcore. I was like, “Okay, I have to talk with every partner after I go contact every company.” I was so hyped up about it and other people were more calm, let’s settle down. But no, I mean, it’s really good. They give you high quality advice. We always get nuggets of insight from our advisors.

When we started a Relationship Hero, we’re actually focused only on text chat. One thing our partners told us is like, “Hey, have them give you a call.” On the way to a YC group advising session, my co-founder, Lior was on the phone giving coaching to one of our clients. They give you great advice. Obviously, the straightforward application process and the straightforward 125k that they give you is nice. The stamp of approval is nice. I mean, YC is killing it. They’re number one.

It’s so fascinating to see kind of just success after success. You wonder how much of that is the selection bias versus the actual value that they derive. What would you say was the best part of the experience and then the worst part of the experience?

Okay, the best part of the experience, I would just have to say how good they are at helping you raise more money. It’s really amazing what they do. I think everybody knows, right? You come in at a valuation that’s less than 2 million, right? When they give you money, you’re valued less than 2 million. They got 7% of the company. By the time you’re pitching on demo day, we raised that 6 million. These days, companies are raising at 10, 12, 15. Twenty is not unheard of. It’s kind of like a 10X in three months, right? Talk about a 10x return.

Yeah, that magical process worked on us to some degree. I mean, it was still hard work to meet with a bunch of investors and raise money but they absolutely give us a boost. That’s probably, if I had to rank the reasons to do YC, I just think the fact that they opened doors to talk with so many investors is probably the number one value.

The worst part of YC, I guess, I might have to say randomly when you have to go listen to the talks, because sometimes the talks are a really special part of YC. I really liked seeing Joe Gebbia from Airbnb. He talked to us and that was really cool. Drew from Dropbox but some of the talks were like we were less excited about and we even skipped on because we were an hour away when YC was [inaudible 00:23:16] and we were in San Francisco. Then, they’re like, “Oh, why don’t you come into the talks?” I’m like, “Well, I watched so many startup talks. I don’t want to see every talk. I kind of opted out of some of the talks.”

That’s awesome. You’ve got a blog you’ve worked on for a while, Bloated MVP. What was the thought process behind that?

Oh, yeah, Bloated MVP. I like to write but I’ve never found until now a theme that I can write a bunch of articles about. Bloated MVP is the first time when I’m like, “I bet I can write 50 articles about bloated MVPs.” The concept, obviously, MVP stands for Minimum Viable Product. Bloated MVP is kind of like another synonym for a startup that’s not a lean startup, right? If you try to follow Eric Ries’s lean startup and you try to make an MVP, but you fail at it, you end up with a bloated MVP.

The reason why I want to write about Bloated MVPs is I think even today when everybody knows you’re supposed to make an MVP, right? It’s like it’s already seeping the culture and yet 80% of companies that I see are Bloated VPS. I’m like, “What’s going on? Why is everybody still shooting themselves in the foot?” I made a blog, it’s bloatedmvp.com and it just documents like, “Okay, here’s another startup. Shooting yourself in the foot. Making a Bloated MVP.” It’s so easy not to make a Bloated MVP if you follow a few principles. That’s basically what I spell out.

Why is it so easy to mess it up? Why does everyone continue to launch these Bloated MVPs?

Yeah, there’s actually a bunch of reasons because you think the concept is so well taught. You think everybody would get it by now. There’s a bunch of factors that lead you astray. For example, even the word MVP, Minimum Viable Product, the word product I think is actually very misleading. Remember how I said Relationship Hero? It started out as me and my cofounder just helping five of our friends? Okay, how do we help five of our friends? They would text us screenshots or we made a Facebook group and we post it in the Facebook group. Would you call that a product? No, but it was us giving value to our friends. I would still call it a value transaction.

What people don’t understand is, when you want to build a company from nothing, you have to follow the value. You have to ask yourself, “How do I create a little bit of value? How do I double the value? How do I create more value?” When you do all that, that’s not necessarily a product but what most founders do is they’re kind of, they’re cargo coating, right?

They’re copying the surface level characteristics and like, “Okay, here’s my startup. I got to get a cofounder. I got to get an accelerator. I got to get a VC Angel. I got to check all these boxes to do a start. Then, I got to hire a developer or learn to code, right? Then, I have to build a website launching the app store and go to product time,” and everybody’s just like, they’re like zombies or like lemmings, right? It’s like, “No, forget all that stuff. Just follow the value.”

Yeah. It’s crazy, too, because I think that it’s so easy to get lost in the weeds and want to build feature after feature after feature but also, there’s a psychological element to it, right? Yes, you want to make the product as good as you can but it’s also scary to launch something that you know isn’t good to some degree.

Yes.

How do you think you have to get over that?

Yeah, definitely. There’s a few tips. One of them is you’re like nobody cares. I mean, there are certain high profile launches that naturally hype up, right. If a well-known entrepreneur is going to go launch something, I actually think, Marissa Mayer launched something the other day and she’s so high profile that she would probably have to launch it anonymously if she wanted to get some early feedback and iterate but for 99% of entrepreneurs, including myself, people care so little about what you launch. They just care so little. It’s like people imagine that they’re going to launch and they’re going to get a million people, like the intense spotlight on them. There’s not going to be a spotlight on you. You should be so lucky that a bunch of people will care to judge your product. Just launch it and see what happens.

Yeah, and the downside really isn’t that bad because there’s so little people using it that if it absolutely sucks, everyone hates it, you’ll learn and you know, what do they say, feedback is a gift, right?

Mmm-hmm

I want to talk a little bit about modern rationality. This fascinates me. You’ve been kind enough to suggest a couple of guests to come on the podcast previously to touch on this stuff. Just let’s start at a high level, what is modern rationality?

Yeah, sure. Rationality is really awesome. It’s fun to talk about because there’s so many misperceptions about it. I have to give credit to where I learned all the concepts of modern rationality, probably centered around that site, I mentioned lesswrong.com but there’s also overcomingbias.com, which is Robin Hanson’s personal blog and yeah, he was on the podcast. I highly recommend listening to his episode and lesswrong.com was actually founded by Eliezer Yudkowsky, who is somebody, he’s very prominent in the artificial intelligence research community and one of the inventors of modern rationality.

He wrote this thing called the Sequences, which is like many thousands of pages of dense explanatory content have all these different ways of how to think and how to properly change your mind like exactly when the evidence tells you to change your mind. Basically, like how to do science better, very high level, what is all the stuff, right? What’s the central idea? It’s basically like how to wield your brain as a tool because obviously, your brain is extremely powerful, right? It’s the most intelligent object in the known universe but at the same time, it did just evolve, right? It wasn’t designed. It evolved and the way that it evolved was to survive and reproduce.

Your brain is not designed to see truth and have accurate beliefs. It’s just designed to let you get by in society. Rationality is about trying to remove all the filters and just be like, “Well, what if I wanted to use my brain as a tool for clearly seeing the world and making discoveries? How do I do that better?

As you do this and kind of learn this way of thinking, is it practice makes perfect and you just got to really kind of throw yourself off the deep end and just obsess over as much of the content as possible and you pick up kind of muscle memory or is it something where somebody can sit down, read for five hours, get it and then they’re off to the races?

It’s a good question. How does one actually become more rational and one of the principles of modern rationality is that a lot of times it helps to focus on a problem. Don’t just focus on the rationality itself, but focus like you’re trying to solve a problem. For example, if you’re trying to go to the moon, that necessarily means you have to get rational about how orbits work. You know what I’m saying? You have to kind of chain backwards and you also have to get rational about, you have to think, how do I design a protocol to test different iterations of my rocket engine? You see what I’m saying?

Not only do you have to understand the physics of the engine, but you have to step back a level and you have to think how do I organize a bunch of humans to build a rocket engine and when it doesn’t work, how do I tell them to evaluate the data and try again, and when should we do the experiment? Should we put a bunch of parts together and test them all at once? It keeps making you step back and rationality is kind of like the ultimate stepping back of how do I organize everything to begin with?

Yeah, and it feels also like there’s a mental framework that you can apply to almost every situation but really, this depends on this intellectual curiosity to continue to drive towards whether it’s first principles thinking or really getting at the root of a problem. It almost feels like what you’re doing is you’re trying to stoke or build up that intellectual curiosity muscle, while also developing these frameworks in which to apply it so that when you’re looking at a problem, you’re not kind of getting caught by all the shiny things and you’re really finding the meat of the issue?

Mm-hmm (affirmative). Yes, there’s really, it’s a whole toolbox of frameworks. There’s so many greatest hits of rationality. One of them is just the idea of changing your mind is good. Imagine two people go into an argument, a lot of times both people are like, “Well, this is what I believe and this is what he believes,” and we’re both going to try to convince each other.

One of the things of rationality is the idea of like, “Well, if you’re really just trying to have a more accurate belief, you should think of it like a dance, almost like a martial art,” where you come in and inside of your brain is a probability distribution over possible ways that you think that the world might be. As you interact with somebody else, the stuff they say, you take it in, and you want that stuff to update your mental model. You want to let your mind be changed. If your current mental model is not accurate, you want to let it get more accurate by changing it, right? Not every change is going to be an improvement, but every improvement is going to be a change.

That’s the kind of stuff rationality tells you. In order to pull that off, they have countless bags of tricks that help you do that. I’ll give you one that I think is really profound for me, which is just the tool of specificity, which is just like when you hear an argument, before you get really reactive to the argument, try to zoom in and drill down and make it more specific.

I’ve written about an example on my blog. Imagine that like, I’m somebody who’s pro-capitalism, and I think Uber is like an overall good company, but I know it’s kind of controversial. Somebody will come up and tell me, “You know what I don’t like about Uber? Uber exploits its workers.” I’m tempted to say, “Well, in capitalism, anytime you’re giving somebody an offer of employment, you’re just making a bid in a market and so they can take or leave the bid, but overall, it’s driving wages up, right? Because they’re bidding for labor and they can always get a different…”

Anyway, that’s my temptation is to give this high level answer but here’s a rationality trick I learned, before I even give any answer, I can stop and be like, “Okay, Uber exploits its workers? Can we get more specific about that? Can you tell me a hypothetical story of one person and just describe in detail how they’re getting exploited?” They say, “Okay, well imagine a guy named Joe who wants to earn a few bucks. He takes a job at Uber and he ends up making $13 an hour and he drives around and his car gets some depreciation and he doesn’t get benefits.” I’m like, “Okay.”

The guy’s like, “Well, and meanwhile, Uber’s IPO and they’re making 80 billion and Joe’s like making $13 minus depreciation.” I’m like, “Okay, well, is that really exploitation? What’s Joe’s next best alternative?” “Well, Joe could go get an actual job with benefits.” “Okay, then, but why didn’t he?” “I don’t know. I didn’t know.” “Well, you have to tell me why he didn’t because you’re telling a story. I’m just asking you to tell me a specific story. You can go ahead and flush out the detail. I’m just trying to listen.”

When they try to flush it out, they’re like, “Well, maybe Burger King offered them benefits and Uber’s not offering them benefits.” I’m like, “Okay, is that why they’re exploiting they’re not offering benefits?” I’m not saying I’m going to win the argument but I’m saying just by being really specific about what they’re even talking about, a lot of times, they’re not talking about anything coherent and their argument just kind of like collapses.

I love the idea of just keep asking why and how and why and how and really, really pulling out the individual details.

Right.

One area where, I don’t think we are yet able to do that, is super artificial intelligence. You’ve told me that that is a big topic among this rational community. Help me understand kind of what is the obsession there and kind of what are those conversations about?

Yeah, totally. Really important topic and it’s actually, I mentioned that Eliezer Yudkowsky, the creator of lesswrong.com is somebody who’s big in both the AI community even the rationality community. His initial motivation for teaching the world about rationality is because he realized that he wasn’t getting enough traction educating the world about artificial intelligence risk, because he needed to first step back and be like, “Listen, this is how you need to use your brain. Okay, now that your brain is fixed. Let me tell you about AI risk.” That was like his course of events.

The argument for AI risk, luckily, it’s been getting a lot of traction. If you compare today, if I say, “Hey, AI is risky. We need to study AI risk,” a lot of people nod their head. Like, yes, it is kind of risky, yeah. If I said this 10 years ago, it would just kind of sound like a very random fringe topic to bring up. It’s definitely becoming more mainstream. Elon Musk has said, “Yes, we need to study AI risk,” but what is exactly the risk?

Very simply, it’s just the idea that it’s very possible to imagine an AI that’s really smart and doesn’t mind accidentally destroying a bunch of stuff, like destroying life as we know it. It’s kind of easy to accidentally destroy life as we know it. For example, imagine if you took like a random animal, like a badger and you just scaled up the badger until its capabilities are more than a nation state? Well, you can imagine the badger might accidentally destroy the world if it doesn’t consciously try not to. That’s kind of like super intelligent AI, similar kind of risk.

Yeah. It’s fascinating in terms of, there’s a line between, “Hey, we need this. It’s going to make us so much better at XYZ activity,” and then there’s a quick turn over that hill and then it becomes, “Hey, we should really fear this thing. I could kill us all.” Where do you think that line is and how much of a validity do you put in the idea of super intelligent AI risk?

Yeah, I actually think it’s a really, really important big risk. Remember, we were talking about like the 10% chance of 100x return and I was talking about order of magnitude perspective, your ability to notice when something is like a big deal. AI risk is a very, very big deal. The chance of it is actually a lot more than 10%. The return of getting it right versus wrong is like a lot more than 100x. It’s a very, very, very big deal.

People are just, they’re having a hard time intuitively getting it because when you look at AI today, it’s like, okay, a classified image. People say all these smug comments, like, “Oh, alpha go can play go,” but if I were facing it in real life, I could take out a gun and shoot it, right? Or I could turn it off, I think Neil deGrasse Tyson, I’ll find the off switch or whatever. But it’s like, no, it just it doesn’t work like that. It’s so easy conceptually for the AI to be like, “Well, I’m going to manufacture a nanobot by Sunday.” There’s labs right now that are connected to the internet that have enough tooling to start.

There’s biology labs connected to the internet right now and you saw how today we have the Moderna vaccine, right? It’s mRNA. They just coded some mRNA and you put it into your body, and it basically programs your body. Okay, well, if an AI were sufficiently smart, it would just have to be connected to the internet to start doing bioengineering like you could pay with Bitcoin, you can build these things. The problem with nanobots is this is just one thing they can do. I’m not saying they will. I’m just giving you a possibility, like nanobots, the problem is they can replicate a bunch and you can end up with the entire Earth turned into nanobots pretty quickly because of exponential growth. I’m just saying there’s a lot of power here that you can unlock if the right code is running on the internet.

Yeah, and what’s crazy to me is like everything you just said, there’s 90% of people or more will be like, “That’s crazy. There’s no way nanobots are going to be running around doing all this stuff,” or whatever but actually, no, there is a possibility. I always push people on the probability and then the asymmetric payoff. Bitcoin is, a great example. Five years ago, probability of it succeeding still pretty low, low single digits, probably but very, very high asymmetric payoff, if it works. It’ll be very big, kind of that order of magnitude that you’re talking about.

This is even bigger and worse, right? The probability of that happening is probably even lower, which means that the asymmetric opportunity is even more egregious and therefore, I guess that’s really where you’re kind of you drive like, “Hey, people should really be paying attention to this, even though it’s a low probability because the risk is so big.”

Well, funny enough, I actually think it’s pretty high probability too.

Why do you think that?

There’s a few things. One is just the idea of the… Okay, first, let’s start with what are the chances that computers are going to get smarter than humans? I think that’s like 80 or 90%, right? It’s hard for me to imagine that 100 or 1000 years from now, the human brain is still going to be the smartest thing, that doesn’t sound plausible to me because computers are rapidly getting more intelligent and capable.

Yeah, well, I guess, really, that it comes like timeframe, right? It’s like, if there’s a larger probability, it’s just on what timeframe? In our lifetime, is that in the next 200 years or is that 10,000 years?

Yeah. I think it’s probably in our lifetime. I would say, it’s within like a few years to one or two lifetimes. I would be pretty shocked if 200 years from now, we still don’t have smarter than human intelligence. I’d be like, “Really? What happened?”

Do you think that means that we shouldn’t develop super intelligent AI? Should we just stop?

There’s a good tactical argument to say we should just stop but it’s also the problem with stopping requires a ton of coordination because it’s not… Back in when they’re doing the Manhattan Project in the ’40s, you could argue, you kind of needed a nation state to enrich uranium or whatever. You could just be like, “Okay, let’s ban enriching uranium and same as we try to do with arms control today,” but when it comes to making an AI, it could literally just be like a team of less than 10 really smart people in a basement with a computer. There’s no way to force them to stop even if the law says to stop.

Yeah, it’s pretty crazy. When you think back to rationality, are you able to apply these concepts on a daily basis or is this more like a framework that you just use as you kind of go through the higher level decisions in your life?

I personally find that I’m able to apply it quite a lot. Just probably my favorite thing to apply besides what I said about making people get specific, besides that one, my other favorite thing to apply is just throw out a numerical guess even if it’s vague as hell. With Bitcoin, Bitcoin back in 2011, when most people see that, they’d be like, “Yeah, it has a very low chance,” and then they would leave it at that but what I did was I said, “Okay, well, is the chance more like 0.001%? Is it more like 1%? Is it more like 10%? Is it more like 50%? I said, “It’s about 10%.” Where did I get that number, right? Out of my rear. Where did I get 10%? But I just think 10% is a better guess than 1% or 50%.

Making that guess it lets you make another guess. What is the upside? Is the maximum possible upside? Is it 10x? 100x? 1000x? I’m like, “Okay, let’s say 1000 or 10,000x.” Then, what I have to do is just multiply 10% by let’s say, 1000x, suddenly I get this number 100x and like, “Wow, okay, well, the expected value is 100x. Sounds like a good investment to put a little bit of cash into. That’s useful thinking.

I mentioned that in the concept of investing, but you can even do it in the context of work, should I do this project? Can this project 10x my company if it works? What’s the chance of it working? Can at 1.2? Can it grow my company 20%? I’m always making these vague numerical calculations and I find that they’re more useful to do that than not.

Yeah. Let’s talk about Relationship Hero to finish up. As you built the company, obviously, you’ve got this framework of rationality. You’ve been early to a lot of different technology trends. Where do you think Relationship Hero is going and kind of how do you think about the path to get to that kind of final vision?

Yeah, totally. I think, the big vision of Relationship Hero is there just needs to be a really big relationship coaching industry. It needs to be like a 10 to $50 billion a year industry. Funny enough, you can look to our neighbor therapy, right? Therapy already is a 30 million industry in the US or a $7 million industry because 30 million Americans go to therapy. You already have 10% of the population going to therapy.

Ask yourself, what percent of the population would benefit from a relationship coach? Well imagine anybody you talk to ask them to make a pie chart being, in this pie chart, make a pie chart of all the things you care about in life and how much value they are for your whole life? A big slice of the pie is going to be the relationships, right? Suddenly, you’ve got 300 million Americans, they all care about the relationships. Where do they go for relationship help? Instead of going to their friends, they should go to a relationship coach. The vision for Relationship Hero is to just take what we’ve been doing now for 50,000 clients and scale it up to 50 million clients.

Yeah. In terms of how big it can be, do you think it can be bigger than the $7 billion kind of therapy market or do you think that it’s probably about the same size?

Well, I think that it’s within an order of magnitude, right? I don’t think it’s going to be more than 10 times bigger but I also think that it’s not going to be more than 10 times smaller. I think it’s going to be maybe it’ll be like 50% bigger. I think, a lot of people need therapy and therapy is great but for me, if I had to guess, I would actually think that more people could use a little bit of relationship help compared to a therapist, but I don’t think that they need to be compared because I think they’re both good and they’re the same order of magnitude but if I had to pick one, I’m going to say a relationship coach.

For sure. Somebody who uses the product today, what is the experience like from a user’s perspective?

We make it as convenient as possible. If you want to do a video call with a relationship coach, if you just want to text them, if you want to get on the phone, you pick, you just go to relationship.hero.com. Actually, since you’re watching the Pomp Podcast, you can go to relationshiphero.com/pomp and we’ll give you $50 off. I got to do the promo. Then, you just schedule an appointment. We do same day appointments. Oftentimes, within an hour, you could be talking and you can just say your issue. We have a money-back guarantee.

The reason why we’re able to run this business and have it be profitable is because we have an extremely high satisfaction rate where a lot of times people come in feeling stuck but then they report to us like, “Oh, wow, you really opened my eyes. This coaching was surprisingly valuable for the time and money I spent.” That’s why we think we’re onto something.

Yeah. What’s so fascinating about this, I think, is not only one is there, that money back guarantee, you’ve put thousands of people through it. The discount, obviously relationshiphero.com/pomp, which everyone appreciates, but also there’s a very quick feedback loop, right? Sometimes people say like, “Hey, come and use my product or service,” and like, “Six months from now, you’ll know if it works or not.” That’s pretty hard to sell. For you guys, a lot of times, it’s literally within hours or days somebody will know immediately either this is helpful to me or not. That high satisfaction kind of rating really drives home the point that this is working for thousands and thousands of people.

Yes. It’s very interesting about relationship advice because it’s very hard to give the right advice but when you hear that you’re getting good advice, a lot of times even without being an expert, you can feel that you’re getting good advice. When you hear it, you’re like, “Oh, yeah, that actually makes a lot of sense, even though I didn’t think about it.” Even within the same session, you’re already getting some feedback from your brain being like, “This is making sense.”

I love it. Before I let you go, I ask everyone the same two questions. First is what is the most important book that you’ve ever read?

Well, I have to say the Consolidated Sequences from LessWrong. Actually, if you go to lesswrong.come now, they just, for the first time ever, published the LessWrong Sequences as a book and they published a collection from last year’s LessWrong Essays. I guess I’m going to promote LessWrong’s book because it really is. I’ve read everything three times over and we’re talking thousands of pages. I want to put in a plug for them.

They put it as a book that you can just buy?

Exactly, yeah. It’s like a physical book. It’s a Kindle book. Yeah, lesswrong.com.

All right. People should go check that out. Second question, aliens, are you a believer or nonbeliever?

Man, I heard you asked us to Robin Hanson and he gave such a good thorough answer. I’m not going to top Robin Hanson’s answer but I’ll give my like dumber take, I guess, which is I think that in the observable universe, it’s pretty obvious that there’s no alien activity, just because if they were really smart and serious, they would obviously be harvesting all the resources. They wouldn’t leave the stars out there to just burn themselves out. That’s like a massive waste of resources, for example.

I think that if there’s probably aliens right outside us in terms of like computer simulation levels. I’m definitely and you might want to ask these question instead of just asking, “Hey, are there aliens?” Are we in a simulation? Because if I had to guess, gun to my head, are we in a simulation, yes or no? I think I’d say yes just because think about this, we’re in an era where we’re inventing virtual reality, right? I can basically put on virtual reality goggles and feel like I’m in a simulation. It’s like we’re inventing simulations. This technology is just becoming possible. We’re going to start like, there’s going to be people who spend their whole lives in VR and yet, you’re telling me there’s no other aliens who have invented a simulation and put us in it? We just happened to be in a time when simulations are getting invented? It seems like too big of a coincidence to me.

One of the craziest parts of me asking the question about aliens is one, how many people believe aliens are real? 90 plus percent, probably maybe 95% of people who come on it to the point where it’s shocking to me when someone says no now, that completely inverts when you ask about simulations. It’s like less than 5% of people think that simulation is real.

What I wonder is how much of it has to do with the socialization and the political correctness of these ideas, right? Aliens 20 years ago, like you were the crazy person if you talked about aliens. Now, it’s like, did the government has UFO videos and all this kind of stuff. Now, aliens have become much more acceptable to talk about and believe in publicly. Simulation, my guess is 10, 20 years from now will be very similar. It’s just at the moment, not quite as popular of an idea to kind of embrace as maybe aliens are.

Yeah, Overton window has embraced aliens but not quite simulation.

Absolutely. You could ask me one question to finish up. What you got for me?

Well, Pomp, I’ve kind of watched you build up your podcast and your audience and it’s very impressive. My question is, what’s been your most effective strategy for growing your audience?

Consistency, not even close. It’s literally just doing it over and over and over and over again, day in day out for years. That sounds like too simplistic for most people but when I first started, a lot of people forget, we didn’t have video on YouTube. We literally were really, really bad at the audio stuff. We didn’t know what we were doing. The guests kind of were all over the place in terms of some like really high quality guests, some that just, somebody emailed me and was like, “Hey, I want to come on your podcast.” I was like, sweet. That must mean it’s legit, bring somebody.

But as we’ve improved, I think the quality, the guest, all that kind of stuff what I realize is like, a lot of the most diehard loyal listeners and readers and things like that for everything that I do, came from the early days. It didn’t matter that it wasn’t high quality in terms of production and all that. It was just the fact that I kept putting out the content and they got value out of it and that’s why they stayed but I think the consistency, by far, is the number one thing that I’ve learned. That’s always my advice to other people is like, “Hey, if you want to do this, first piece of advice, don’t do it because you literally have to be consistent for years and years and years.” I think I’ve been doing it now since the end, 17, 18, end of the fourth year.

It’s just when you think about how many people are willing to sit down and just bash the head against the wall for four years or so, not many. A lot of people think they are but when you get today 472 and you have to go record a podcast and write and do all this kind of stuff, I think people just say, “Hey, look there’s better stuff for me to go do.” I think consistency is definitely the differentiator.

Yeah, yeah, you’re very prolific. You’re definitely one of the top podcasters in my feed in terms of just like constantly releasing episodes.

I try five times a week. That’s the goal. We’ll see if it works.

Oh, yeah. Well, congratulations on getting this guy right here, top talent.

Exactly. All right. Where can we send people to find you on the internet and find more about Relationship Hero?

Cool. Yeah. I’m going to say relationshiphero.com and check out my blog, bloatedmvp.com.

All right, bloatedmvp.com and then relationshiphero.com. If you use .com/pomp, what is it, $50 off?

$50 off.

All right, guys, go check it out. Thank you so much for doing this. Let’s do it again in the future.

Thanks a lot, Pomp. I appreciate it.