Steve Hsu: Welcome to Manifold. Today my guest is Stephen Grugett, the co-founder of Manifold Markets. You heard that right, they ripped the name off from this podcast. It's such a good name. But, it's okay. Manifold Markets bills itself as the largest prediction market in the world.

Stephen is going to explain to us how the market works, why he's doing it, what they hope to accomplish. But before we get into that, I want to delve into his background a little bit.

So Stephen, tell me, where did you grow up?

Stephen Grugett: I grew up in Atlanta, with my brother, who is now my co-founder.

Steve Hsu: And, what were your interests when you were a kid?

Stephen Grugett: Ah, a lot of things. I spent a lot of time reading, all sorts of stuff. I was very interested in philosophy and languages. I guess I got some, like, proto economic and trading experience playing games like, RuneScape and Neopets, where the, you know, buying and selling and trading aspect of the game features pretty prominently.

Steve Hsu: And I think I was connected to you through people that knew you in Austin. Did you go to UT Austin as an undergrad?

Stephen Grugett: No, I didn't. I actually lived in Austin for a couple years during the pandemic. I'm a California refugee.

Steve Hsu: I see. So yeah, so let me trace your course. Grew up in Georgia, in Atlanta, and then how did you get to Austin?

Stephen Grugett: through a very, very long and circuitous route. I was an undergrad at Yale. I studied comp sci and philosophy. Then I worked for SIG, an options trading firm, for a year as a programmer. Then I moved to New York for a year. I worked for my friend's robo advisor startup.

Then I sold all of my possessions and became a digital nomad, living around the world, mostly in Southeast Asia.

And during that time I started my last startup with my brother James, which was a subscription group chat app for online creators. Basically like only fans for not porn for every other DJs. Wow. Content creators that sort of thing.

Steve Hsu: You know, I didn't know you were a digital nomad. One, one of my fantasies is to be a digital nomad and live in Southeast Asia? Where did you live?

Stephen Grugett: A bunch of places. I probably like the plurality of my time spent in Malaysia, mostly KL. Yeah the capital. Yep, Kuala Lumpur And, in Bali. And there are a bunch of, you know, digital nomads doing the exact same thing.

Steve Hsu: Yeah, it looks great. I don't know if my wife would agree to this, but I keep joking with her that, you know, when we retire, we'll go live somewhere like KL or Da Nang or Penang or one of these places.

We went to Bali on our honeymoon, which was actually a little disappointing.

Stephen Grugett: Which parts did you go to?

Steve Hsu: Well, the problem, it was, it's beautiful. The beauty is as advertised, but the population density is actually kind of high. And, we were unprepared for the beggars. It really was a big downer to go to a beautiful temple surrounded by terraced rice fields, rice paddies, and then just have like some crippled person trying to get money from you all the time, and so that part of it I didn't like.

Stephen Grugett: I see. I think that may have gotten a little bit better in the year since.

Steve Hsu:Yeah, that's great. And when were you at Yale?

Stephen Grugett: I was class of 2013.

Steve Hsu: Oh, okay. So long after I was there. I was a professor there in the mid to late 90s. Oh. So long before you were there. Okay, great.

So we know a little bit about your background, and we are conducting this interview at the Manifold Markets World Headquarters. Is it a secret location?

Stephen Grugett: It's not a secret location. It's on Divisadero near Alamo Square in the heart of San Francisco.

Steve Hsu: Is this considered Hayes Valley still or are we outside Hayes Valley?

Stephen Grugett: Maybe not technically in the narrowest sense, but Hayes Valley is kind of a state of mind. We're in the extended Hayes Valley.

Steve Hsu: You're Hayes Valley adjacent. Yeah, exactly. So I'm staying close by, and part of my visit here is to mind meld with some of the AI leaders. And for those of you who don't live in San Francisco, Hayes Valley is often jokingly referred to as Bayes Valley, as in Bayes Theorem, because people here are rationalists, A. I. enthusiasts, etc.

Stephen Grugett: Or the even cringier Cerebral Valley.

Steve Hsu: Yes, now I, yeah, that one I don't get. Like, where, who came up with that?

Stephen Grugett: This was a journalist who, Took someone's beat on Twitter at face value. Mm hmm. No one and then it just became a meme.

Steve Hsu: I see. It was a gag and then the journalist picked it up. I like Bayes Valley better.

Stephen Grugett: Yeah, that makes sense. Yeah.

Steve Hsu: It turns out one of our I don't think that's exactly how I know you but one of the AI engineers at my startup Super Focus, when we first hired him, he was in Austin, but he soon moved to San Francisco and he lives around here in Hayes Valley. So, this is indeed the mecca for AI.

Okay, so let's talk about manifold markets. Now it's not very old, right? When was it established?

Stephen Grugett: We're about a couple years old at this point.

Steve Hsu: And was it in fact established with funds that you got through Scott Alexander?

Stephen Grugett: That's kind of the, the thing which, you know, kicked off Manifold as, as a business. Originally, we were just a project. It wasn't clear what exactly this would become.

Steve Hsu: And, so, now you're, I was kind of amazed because when I came by your office I was expecting it to just be a few people. But I just took a tour and, what's the headcount here? There must be like a dozen people working here, is that right?

Stephen Grugett: So not everyone who works in the office is an employee of Manifold. We let a few people here rent out hot desks. Got it. Got it. Our headcount hasn't bloated that much. We're, we're about eight full time employees and, three, part time.

Steve Hsu: Okay. Great.

And, so let's, let's explain Manifold Markets to the listener because not everybody's familiar with it.

So first we'll start with what is a prediction market and then we'll talk about what. special aspects. What are special aspects of the manifold market? So the prediction market is a place where more or less you bet on outcomes. Is that right? Exactly. And the price of a contract, I've not studied the actual system they use at manifold markets, but oftentimes the price of a contract is meant to represent the implied probability occurring.

Is that how you guys do your pricing?

Stephen Grugett: Exactly. Yeah. The basic premise is typically you have, you have a well defined contract, which is usually on, a precisely defined question like will Biden win the, the upcoming presidential election and what people are buying and selling our shares in the outcome of that event.

So for instance, if you purchase a share, That in yes, that is, if the event happens, that share will pay out one dollar, or one mana, in our case, that's our internal currency, if the event does happen. if you're paying 50 cents per share, then, you know, the implied market odds of that event are around 50%.

Steve Hsu: Got it. And, and the original, so, so I have to confess that I've been a friend of Robin Hanson's since, gosh, I want to say Early two thousands, maybe around 2005 ish. And the very first time I met him, it was a conference at HP Labs, their research laboratory. And, I forgot what the name of the conference was, but I was speaking about encryption, related techno internet technologies.

And he was speaking about prediction markets. And that was around the time when I think he wrote the first papers, working out the math for these things. So I've kept in touch with him over the years about prediction markets. I always thought it was a great idea. In his case, I think he tried, he got several big companies to instantiate.

Internal prediction markets ostensibly to uncover information that maybe the rank and file of the company understood, maybe because of proximity to customers or proximity to the products that the company was developing, et cetera. and then would. try to surface that information, for example, to the leadership of the company.

But I think Robin told me that the leaders didn't actually want that information to be publicly visible because it sort of constrained their decision making capabilities too much. And so ultimately, I think Robin told me about a decade later that he didn't feel prediction markets in that context were a success.

Now, I think your goal is something different, so maybe, we'll get into the details of how your market runs in a second, but what do you, what is your goal for creating this marketplace?

Stephen Grugett: Really it's to bring the benefits of, you know, explicit quantitative forecasting to the world by allowing people to predict on and bet literally everything.

Steve Hsu: Right. And so you want to collect information from, in your ideal world, there would be a very large pool of people using your platform and through their using the platform, they would surface better calibrated estimates of the probability of future events. Yes, exactly. So you're in a way, you're providing a public good because everybody would like to know, like, suppose I read something about, you know, the US going to war with Iran because of some cargo ship being sunk or something.

And, I'm, I don't have time to go and read 50 articles about it. Most of which are, you know, lies by the mainstream media, but, but I don't have time to do that. So I just manifold, marketplace and see what the collective wisdom is, the price of that contract. And it saves me a bunch of time. And the people who are, who really have, expert, knowledge about that event or the likelihood of, in this case, war with Iran, they would be in there, they would be incentivized to trade because they have an advantage over other, players in the market, and they would drive the price toward the true probability, that the event would occur.

So you, in a way, then you have a machine which predicts the future as well as humanity can predict the future. Right, exactly. Excellent. So, that's your goal and everyone would love to see something like this instantiated. It would be awesome if everybody participated in it. now the incentive we're talking about here, in your case is not real dollars, as I think you just mentioned a moment ago, but mana, so the unit, the unit of currency on manifold markets is cleverly called mana and you buy in When you join the site, you have to pay?

Stephen Grugett: Well, we give you, it's similar to a free-to-play game. When you sign up, we give you some mana. If you want, that's the last mana that you can ever earn. You know, we have additional ways of earning mana, too, without purchasing. If you want, you never have to pay into the system at all, and you can use Manifold, like, purely as a game.

If you want, though, to subsidize your own questions, or create additional markets — or you just like to be able to bet larger amounts, then you can purchase mana and USD.

Steve Hsu: Got it. And, the reason for structuring it that way, I believe, is legal issues, right? I mean, if you, if, if, if it were just pure dollars being transacted, then you could be considered by the U.S. government is basically a kind of casino. Is that right?

Stephen Grugett: Yeah, that is correct. I would say primary, that's the primary but not exclusive reason. So when I say play money people often think that it's merely an inferior version of real money, but that's not the case Play money actually is its own separate thing that opens up whole new avenues That wouldn't be possible in the real money world for instance like on our platform We were able to have lots of questions on things both silly and serious that we simply wouldn't be able to have if there were real money involved.

You know, the reality is even in a world where prediction market regulation became substantially more liberalized, we still wouldn't have the same level of freedom that we do on Manifolds today. And that's a very valuable thing. The other thing is that it changes how you approach the markets in some ways which are very positive.

Many of our users are turned off by the concept of gambling or betting real money. But if you allow them to participate using play money, they treat it essentially as a type of game. And that can actually be a very compelling thing in getting more people on the platform.

And I guess the most important thing about this ecosystem is that it actually works. You know, our calibration results are surprisingly good. Probably much better, much better than what I personally expected, and I'm, you know, I'm an enthusiast and believer in play money, but certainly more so than the average person would expect.

Steve Hsu: Yeah, so let's explore a little bit the pros and cons of play versus real money. And so we can, we can consider just as a thought experiment, a world in which the U. S. government allowed you to do it with real money. And I think like that. Isn't there a market set up by economists at the University of Iowa or political scientists that is actually allowed to use real money?

Stephen Grugett: Yes, that's right. That's one of the first. There have been several other iterations which have used real money. That and PredictIt were given a letter of no action by the CFTC, who is the regulator of prediction markets, to do this in an academic domain. Right.

Steve Hsu: So what, one question I'll ask you later is if in this fantasy world CFTC came to you and said we'll give you a letter of no action, whether that, you would want that or not.

We can discuss that in just a moment. So, so, okay, so, naively I would have said, well, okay, you, your market is going to attract some people who like games. Maybe some intellectuals who like the challenge of proving that they're good predictors, maybe super forecasters, that kind of person. But on the other hand, some people are going to say like, I would love to play in that market.

I actually know a shit ton about the oil industry and I would love to play in that market. But you know what? I'd rather make money with my knowledge about the oil industry trading in the actual financial markets. And so do you feel that the way it currently works, it attracts a certain type of person, but there's some other pools of knowledge that are not sufficiently incentivized to get engaged with you guys.

Stephen Grugett: For sure, and I think that's totally fine. I think it's unrealistic to expect a play money market to have prices as efficient on things that are traded in global equity markets or commodities markets to the tune of like trillions of dollars of notional value. What, what our platform can do is allow you know, close in on the picture of financial completeness by allowing people to bet and speculate on, like, millions of other things that are not currently traded, on global financial markets.

Steve Hsu: Right. and you mentioned the calibration is pretty good in the sense, now, are you, are you getting that by comparing it to other markets where there's overlap in the contracts? Or just, post hoc looking at how accurate the market price is at the end?

Stephen Grugett: Both. so, you know, overall, we track the overall calibration of the platform, to see how well our markets do compared to what their eventual result is.

And that's pretty good. We also have done case studies, where we compare, we can compare, the, you know, predictive results of our markets. With other, other markets, polls, and like, non, non market mechanisms, and we find that they do pretty well, in general. For instance, during the, you know, the last midterm election cycle, our plain money markets actually outperformed, the, the real, the real money betting markets, which is a very interesting and a surprising result.

Steve Hsu: Oh, that's great. Yeah, interesting, yeah.

Stephen Grugett: It's only, it's only one data point though. So the, you know, the, the case against this is that this was one election cycle and for whatever reason, real money markets happened to skew Republican and, fake money markets happened to skew Democrat. And that's just what the results were.

Steve Hsu: I see. also I think, you, you know, you, you were funded by Scott Alexander and, Just from walking around the office, I can kind of guess that you have very deep roots in the Bay Area rationalist community. And so, like, I would imagine, like, also, at least at this point, a lot of, like, your players, your participants are drawn somewhat disproportionately from that community.

Which, in a way, isn't bad, because these are very thoughtful people who try to be rational about everything and analyze things. But, I, yeah, the question is in the long run, Is there anything preventing you, even though it's quite a bit of money, from really reaching your goal of being sort of the best future predicting machine on the planet?

Stephen Grugett: Yeah, you know, part, you know, part of our strategy rather than focusing on the corporate use case is to focus on this like consumer entertainment use case. Yes. You know, you know, we want to make our platform like a fun as well as educational and informative place. And we believe that's our path to mainstream success and adoption.

Steve Hsu: Got it. Of, among the people who, so I remember the prediction markets. Maybe Tetlock ran, the academic guy Tetlock. Ran these for some period of time and there were people who at the end were christened super forecasters. I, I forgot what the criterion was, but, or criteria were, but, those people, some of them actually went out and hung out a shingle and said, you should hire me as a consultant because I'm a validated super forecaster in Tetlock's Platform.

Do you have a similar dynamic happening on your platform?

Stephen Grugett: We do. In fact, we have, you know, if you look at our global leaderboards, you'll recognize a few of those names, from people who actually like the quote unquote registered trademark super forecasters themselves. so there definitely is an overlap between, you know, prediction market trading ability and forecasting ability.

Although that link is probably a little bit less than you would expect. There are some forecasters and super forecasters who have very impressive predictive track records on other platforms, but who are nevertheless not great traders on Manifold.

Steve Hsu: Yeah, this is an interesting point you made to me yesterday, which is that I guess in the other context, we, we just, you just make predictions and then we score you on the predictions.

But here you're, you're betting on your predictions basically, right? And so the amount of capital, amount of money that you, or mana, that you put against a particular prediction, that's a separate judgment call. So in a way there's like a confidence level and a, you know, the actual central value of the prediction, but then there's a confidence level that is related to the actual trading activity on the platform.

Stephen Grugett: That's right, exactly. a lot of trading, or a lot of trading strategies don't necessarily involve or require predictive ability. As long as you have a very good sense of what the current market price is, you can go into other markets. I see. And do things like make markets by posting liquidity.

Or arbitraging between different types of markets, things like that. Following good traders, if you know that they're under betting, you can follow them and bet. You know, none of those are necessarily, or, indicative of forecasting ability. But they are another type of ability.

Steve Hsu: Yes, got it. Okay, so it is a little bit tricky.

So if somebody, you know, somebody became a mana millionaire, and then they said, hey, you should hire me as a professional consultant forecaster because I became a mana millionaire. It might be the case with that person. made all that mana by just being a savvy trader and doesn't really have a particularly good forecasting capability.

Stephen Grugett: True. But that also can be impressive in its own right. Yes. Depending, or you, you also want to look at the track record as well. It's kind of the same, the same problem as evaluating fund managers. Yeah. You don't just care that they made a bunch of money. Yes. You care that they made it in a smart and consistent way.

And if they can show that they did, then, you know, that's reasonable. But there is no definitive sign that they can do so in the future as well.

Steve Hsu: Right. Are all the, for any individual on the, any user on the platform, are, is, is everything transparent? You can see everything that that person has done?

Stephen Grugett: Basically, yes.

So this is one of the other big differentiating features of Manifold. And a thing that works very well in a play money, you know, in, gamified social media context. And would work less well in a real money context. Right.

Steve Hsu: Got it.

So, let's, let's run through some numbers. Like, so how, like, how many users do you have?

How much mana flows through the system in a given period of time? How much money have people bought in, over time? Any, any numbers like that you want to share with us?

Stephen Grugett: Sure. We actually, publicly expose a lot of our stats. If you go to magnaflow. market. com, you can see for yourself. Hopefully we've grown by the time you're listening to this episode.

Okay, just give me the numbers as of now. Yeah, yeah, as of now we have around, around like 10, 000 monthly activities. That's good. Yeah, which is pretty good. so interesting, like the total, measuring the mana supply, you know, the money supply of manifold, like doing that is basically is almost as difficult as the real economy.

Yes. It's precisely what you mean. Is it like the people's cash balances as of the moment? Is it their current portfolio value? But we also provide loans, which have to be paid back to the house. Wait, you have loans? Yeah, we have margin loans for our traders. And you charge interest? they are free, so we, it's, you know, we have very, very stringent leverage limits.

Okay. much lower than typical brokerages.

Steve Hsu: I see. Do you have, so you have some people that are busts that are, that owe you?

Stephen Grugett: Yes, we have had a few busts before.

Steve Hsu: Okay. How much money do you have, how many dollars have you received in exchange for mana?

Stephen Grugett: Yeah, that's a good question. I think, like over a hundred thousand dollars.

Steve Hsu: Wow. Wow. Wow. That's serious. Okay. Well, if you have 10, 000 actives, I guess that you could have like 100, 000 people have ever been on the site, and if each of them gives you a dollar, I suppose that's, you know, that could be 100, 000, right?

Stephen Grugett: Or something like that. It's very front loaded. I see. Or like, you know, in free to play games, the largest buyers are contributing the bulk of that.

I see. And then there's a long tail of people with like 10 purchases. And we see the same pattern on our platform today, too.

Steve Hsu: One of the things I jokingly said to you last night is that, if people really love your game and there starts to be a secondary market for mana, like I go out down the street and buy some mana from the guy in the cafe, then all of a sudden you're basically using a kind of cryptocurrency as your, as your, betting, unit.

Stephen Grugett: Yeah, so this is a problem that other game economies have as well like, the roadblocks, yeah, yeah, well, and typically you, you just like to crack down on, on the exchanges and on the individuals like, engaging in them. And as for us, this hasn't been a problem today on Manifold, but as we scale then we'll probably follow down the same path and do similar things.

Steve Hsu: Now, do you feel like the activity is just a little bit too cerebral? I mean, like there's a probability involved and there's a lot of thought and research. Do you feel like it could only ever appeal to a relatively small fraction of the population?

Stephen Grugett: Well, our goal is to kind of simplify.

You know, the trading interface and the process, you know, the entire process of engaging with a prediction market such that more and more people can use it. I think we've come a good way at doing this and helped broaden the audience substantially. But I think,even though we have further to go, there are a bunch of things that we can do to make it even simpler to participate.

My hope is that everyone will be able to engage with the markets, at the level of sophistication that they're most comfortable with.

Steve Hsu: I think one could argue that people who aren't thoughtful enough to think in terms of probabilities and stuff like that, maybe are, would mostly just be adding noise to the information in the markets anyway.

So, it's the people who kind of like this kind of stuff who you're trying to, trying to capture on the platform.

Stephen Grugett: I guess I have two points here. The first is that, you know, noise traders subsidize the more informed trader, so that's a very nice and beneficial thing. Yes. The second point is that even if you're not a good trader, it actually is very, you know, psychologically beneficial to realize that.

Yes. And to, you know, recalibrate your own abilities.

Steve Hsu: Yeah, that's a good point. So, as of now, I think you mentioned, so, so I imagine some of the contracts that are there are inserted there by your team. But then it sounded like people could commission some, if I were curious about what people thought about a particular issue and you didn't have a contract relevant to that, I could commission you to start trading on a contract, a custom designed contract.

Stephen Grugett: So we have user created markets. We have a self service model. You can just come on our site and create your own, you know, betting market on any question that you care about at all, and then share it with your friends and the broader community.

Steve Hsu: I see. So, okay, so if like a bunch of friends were betting on whether Joe and Jamie are going to get married, we could create our own little internal betting market about that, just for fun.

Stephen Grugett: Exactly, that's a great use case. In fact, my co founder, Austin, Basically, has documented the entire stage of his relationship with his current wife through our markets. I see. There's a market on whether they'll start dating, whether they'll be dating in three months, whether they'll get engaged, whether they'll get married.

Now there's a market on the name of their first child now.

Steve Hsu: Wow. Now if you, let's suppose that though I have a, I would like to create a custom contract. You just said I could do that myself as a user. But suppose I wanted to attract expertise into the trading of that contract so I could try to, you know, gain information about what's going to happen in the world, and bring in some expertise.

Would I then pay you perhaps to advertise the existence of that contract?

Stephen Grugett: Yeah. So we have this method of internally advertising markets on our platform called boosting. So, when you create a market, we require you to put up a really tiny subsidy, which is not that significant just to get the trading started.

But if you, if you want to get more action on your market and, you know, entice the larger traders to participate, we have this thing called boosting, which combines two different things. One is subsidizing the market. It injects some capital into the market to increase the amount that traders can win.

And then secondly, it advertises the market on the feed, to other, other people of similar interests.

Steve Hsu: Okay. But currently that's not a revenue stream for you. Like if you said like. Pay us some dollars and we'll boost the contract that you guys are really interested in.

Stephen Grugett: We actually, we actually do charge a fee on the boosting.

In, in mana. This is all denominated in mana. Oh, you

Steve Hsu: don't take any, okay, so you're not trying to get, you're not trying to get any revenue in dollars from like those side activities?

Stephen Grugett: Indirectly. You know, this is, so this is a thing, One of the primary sources of value for mana is that it's the thing that uniquely allows you to create and like subsidize and promote questions on our platform.

Steve Hsu: I see. Okay. So it's an incentive for people to want to hold mana, to own mana. Right. Yeah. I see.

So what do you think, maybe talk us through what you think, ultimately your revenue streams will be, you know, and where you will go as a business.

Stephen Grugett: Right, that's a great question. So I do think that we'll be able to continue our current revenue model of selling mana for USD indefinitely.

We're getting pretty good monetization rates right now. That's a thing that can continue and there are a bunch of examples of that working in a nearby field. but outside of that, there are a bunch of other options that we're considering as well. There's some very obvious things like having a subscription plan, which would enable, we're thinking about pay gating advanced features like margin loans and other power user things behind, putting that behind a pay gate and charging a subscription.

Then, there's a separate class of things which are more exotic monetization options, which come from the fact that our markets, like, are genuinely, like, predicted. You know, and one of, you know, one thing we're considering is, you know, having, like, a white glove service for, financial firms, governments who, you know, want help in crafting their own markets and promoting them to make sure that they can, get as much predictive power out of them as possible.

You know, that's a service that we could provide. You know, we're, you know, also considering other ways to lean even more heavily into things like this, boosting an advertising use case of making it easier for individuals to get answers to their questions. even farther afield from that, one other idea we've had is, you know, we have leaderboards.

We're actually very good at identifying the top prediction market traders. Those people tend to be very bright. They often tend to be very underemployed. So one obvious opportunity is to just connect them with known employers probably, you know top tier tech firms or financial firms. Yep.

Steve Hsu: I see so you can you can you can become an HR Function you can have a jobs board or HR function actually exactly headhunting headhunting. Now, do you have any venture investors or board members that are pushing you?

You know, what, what's your revenue model? How is this going to turn into a profitable business? A unicorn? Is that a dynamic that's part of your daily life?

Stephen Grugett: Not part of my daily life, thankfully, but our, our investors, you know, our investors are obviously very keen on monetization. We are still a consumer platform, though, at this stage, and early stage, consumer car companies typically are hyper focused on growth.

Yes. And that's kind of where, where we are focused as well.

Steve Hsu: I see. You know, I think the op, my optimistic view of this would be it, it could become normative as a kind of fun activity for brainy people, and if you capture, you know, 1 percent of the population playing this, you know, having a presence on your marketplace.

A, it would be really a fantastic instrument for predicting the future. And yeah, it could be big enough that it actually throws off a fair amount of cash for you guys, right? That's the hope, yeah. So if you get millions of users, that would be, you know, a pretty good outcome, I think.

Stephen Grugett: Yeah, yeah, I think it's readily achievable.

Steve Hsu: Is anybody competing with you with a pretty similar model?

Stephen Grugett: So, no one with our, like, exact spin on prediction markets. We're, we're, play money and we're user created prediction markets. And there aren't any other, you know, firms which have that exact spin. But there are real money prediction markets, you know, as a class of things.

Steve Hsu: And how, how do they avoid the legal issue?

Stephen Grugett: By being illegal. Oh, okay. Well, there are, so there are, there are several classes. So there are, Some, firms which have tried to go entirely offshore and do not, or in theory, have no U.S. based customers, to not run foul of the regulations.

Steve Hsu: But the U.S. based, if you were indeed a U.S. based customer, you could be in trouble if you were.

Stephen Grugett: That's right. Yeah. The CFTC has offered approval for it. Limited types of prediction markets too. So there are firms trying to go for this regulated category. I see. What you see there though is that, these markets aren't the markets people are interested in.

What people want to bet in most are things like elections and geopolitics and things like that. And unfortunately, those categories are prohibited by the CFTC, who is the regulator of this.

Steve Hsu: I see. I see. Oh, the Iowa election markets really got, got a good deal, right? Because they have a dispensation, and they do elections, right?

Stephen Grugett: That's right. But there, the flip side of that is that there are limits. Like, they would not be able, a hedge fund wouldn't be able to come in and place a ten million dollar bet. It's, it's just geared towards, you know, small players.

Steve Hsu: I see.

Now, I was looking at your site and I saw a corner where it said something about dating markets.

What are dating markets?

Stephen Grugett: Yeah, so, this is one of the use cases that our users organically have discovered, by themselves. You know, you know, this is just natural, people's natural curiosity. leads them to ask these questions about their friends and their relationships. You know, whether friend X and friend Y are going to be dating in six months, whether they're going to get married, that sort of thing.

Steve Hsu: So those tend to be basically only of interest to small groups of people among common groups of friends or things like this.

Stephen Grugett: Yeah, although people have a natural instinct to, I don't know, seek out drama and learn about other people and kind of live their relationships vicariously through them.

So, I think what you see is the more you, the more you document whatever the current state of your personal life, the greater engagement you can expect to, to get on that.

Steve Hsu: Interesting, okay, got it.

Now it seems to me you guys are masters of public relations, so you had an event recently, which was covered in the New York Times, like very prominently, wasn't it? Remind me about that.

Stephen Grugett: That's right. Yes, this was our, our conference, Manifest.

Steve Hsu: Yes. And where, where was it?

Stephen Grugett: We held this at Lighthaven, in Berkeley.

Steve Hsu: And, so I read that article and I was like, Damn it, I wish I could have gone to this. Why didn't, why didn't Stephen invite me to this thing?

We didn't know each other then. But, one of the things that caught my eye in the, in the write-up in the Times was that there was a wrestling match at midnight. And so I was thinking, was that, did the reporter get that wrong? Was it actually wrestling or was it jujitsu? What, what, what was that all about?

Stephen Grugett: Unfortunately, I personally did not attend this event, but I believe it was wrestling and did happen in exactly the manner that was described in that article.
Steve Hsu: Interesting. Because I wonder how many, I was very curious because I thought like, how many Bay Area rationalists know the actual rules of either folk style or freestyle or Greco-Roman wrestling?

Because it's a little bit obscure. Like, I come from the Midwest where wrestling is a huge sport, where I grew up. But I was thinking, yeah, it's more likely they're probably doing Jiu-Jitsu or something because, yeah, go ahead. Yeah,

Stephen Grugett: I think that's, I've heard this as, you know, as well, like, Brazilian Jiu Jitsu is now an in thing in the Bay Area.

Steve Hsu: Yes, yes. So I was, I was guessing the reporter didn't know the difference between jiu-jitsu and wrestling, but you're saying it was wrestling? I think

Stephen Grugett: it, I think it actually was, was wrestling nonetheless. I know, I know the person who organized this. Cool. All

Steve Hsu: right. That's great. okay. So anyway, you guys are masters of, you know, getting covered prominently in the New York Times for your conference.

I spent a little time this morning watching a musical that you guys produced. So tell us about this musical.

Stephen Grugett: Yeah, this is kind of a crazy idea. In many ways, this grew out of Manifest as well, where we noticed that a lot of users were having a lot of fun betting on these like personal and relationship markets. Yes, yes.

during this event, and I was trying to figure out a way to, you know, another event to keep the fun going. You know, I immediately, you know, thought of the concept of doing a dating show, which is a great fit with prediction markets. People can predict on the out, on the winner, and it's, it's, it's a lot of fun.

So I had this idea. I originally was going to just host this event myself, and it would have been a very, you know, low key, much lower budget type thing, where I just like, Have one, one camera and a live stream and I, you know, select a few contestants, etc. My employees convinced me that that would be a bad idea and it would be boring and not really entertaining.

so I was trying to shop around and think about ways that it could make it entertaining when one of my friends told me that there had been a dating show at Vibe Camp. So Vibe Camp is like Twitter, highly I don't know, very weird, very niche internet culture, festival conference type thing, which had happened previously over the summer.

Steve Hsu: My friend put me in contact with the, Well, first they sent me the raw footage of this event and I was originally going to spend like five minutes like looking through in the morning and I ended up like watching the entire thing, you know, in the middle of a work day because it was so engaging. So I reached out to the director and at this point I knew after watching, after watching the show, I knew that I had to work with the director and that we could make something really cool and unique and I pitched her on the concept of rewriting the show, but using, making it centered on Manifold and using our prediction markets and she was completely on board, So who, who wrote the songs? Because it wasn't just a dating show, right? It's actually a musical.

Stephen Grugett: Right, yeah. So the scope of the show grew immensely from my original vision.

So the theater company that put this on is called the Classics Department, and they're responsible for all of the cleverness in the show. I provided them. With some background information about Manifold and some specific knowledge. But they wrote everything.

Steve Hsu: But they get nerd culture, you know, they have Oh yeah, yeah, yeah.

Blood Boy and all this other stuff. Exactly.

Stephen Grugett: It wouldn't be possible to produce a show like this otherwise. Yeah. The musicians areTim Bley and Basu. Tim Bley is internet famous for creating the YouTube channel A Capella Science, which you should look up right now. I see. A bunch of, really amazing and beautifully executed, like, science parody videos.

and they, yeah, they, they, they did an amazing job. Originally, this show was not intended to be a musical. But they, you know, created, like, several, really excellent, songs, which the director managed to incorporate into the show. Yeah,

Steve Hsu: I was listening to it, and I thought the compositions were actually really good.

And actually, the guy who was the emcee is incredibly talented, whoever that guy is.

Stephen Grugett: Exactly. Duncan, which is his actual name the, the character that he played, or he used the same name for the character that he plays in the show. I see. He's an incredible talent. Like me, so I saw the dress rehearsal as well.

A large chunk of the content that you saw, during the actual performance was improvised on the spot.

Steve Hsu: I see. wow. Yeah, I mean there's nothing, there's no substitute for that kind of talent. Until AI gets a little bit better. but yeah, the show was great. I think if you guys are into the nerd cult, Silicon Valley nerd culture, definitely watch YouTube.

I think you could do a streaming dating show very cheaply, right? Because you can get the cameras and all that easily. You stream it. you have people betting on the markets while it's on. And, the contestants are probably free. Like if you say, like, hey, whoever wins will pay for your night out. you could probably get lots of nerds and nerdettes.

Stephen Grugett: So, interesting that you say that. So, actually, some of our users have independently organized a couple of these events in London. Oh, wow. and

Steve Hsu: You guys are big in London?

Stephen Grugett: Yes, actually. I haven't checked recently, but for a while London was our top, our single top city globally. I see. I think this is, this would not be true if you included the Bay Area as one as one city.

Steve Hsu: Yeah. Because of my association with a guy called Dominic Cummings, who is a darling of rationalists, probably both in the UK and here, yeah, I can definitely attest there is a similar scene in London of people with kind of similar interests to the, the Bay Area rationalist scene.

And so, yeah, I'm not surprised a lot of people there are interested in manifold markets. So no plans though for the company to produce a dating show?

Stephen Grugett: Probably not. Okay. There's some, some small chance. We'll, we'll see.

Steve Hsu: Okay. Good.

Any other things you want to cover about, manifold markets that I missed, that I didn't ask you about?

Stephen Grugett: Hmm. I guess my, I would turn the tables and ask you, what, what types of things are you most interested in predicting? Where, where do you see, where do you personally see the greatest need, for more, more and better, explicit forecasting tools?

Steve Hsu: Well, I think geopolitics and elections, maybe some economic things, are really the right place, maybe, maybe a little bit for scientific discoveries, but I'm not sure what the utility of that is.

My main problem, actually, for the current situation that we're in in the world, geopolitically, is not the point probabilities for a given outcome. It's actually because the situation that we're in now is that the West has become very polarized against the rise of China and, you know, other BRICS countries.

And so the stories that our media tells in the United States and Europe about what's going on in those countries are largely fabrications. Just, just very huge distortions. Now, people who are otherwise, like, pretty cynical and careful about evaluating their news sources, you know, about other things, I find, like, say, here in the Bay Area or elsewhere in New York, whatever, are very willing to accept, like, a lot of bullshit, you know, about what's actually happening in China or what life is like in China.

Because frankly, very few people have gone there or if you go there, you don't speak Chinese, so you don't really know what's going on. so I think the main issue is just the lack of a deeper understanding of what's really happening in the world. I don't think markets can really cure that. It would be very interesting if you could get participation from people in Russia and China in these markets, because then you, you, you might see some things were like the, the, the view in the Chinese market is very different from the view in the, the Western vision.

Stephen Grugett: I think we're, I think we're actually banned in China, so that's unfortunate. We had some markets like, uh.a year maybe. I don't remember when, but when Pelosi was visiting Taiwan. Yeah. There were a few markets about the events there and for whatever reason they became popular in England.

Steve Hsu: I see. You know, and the thing, the thing with, mainland, like a lot of stuff like Facebook and stuff, is banned in China, but then like people who are zens in China generally have a VPN, so they can actually still go. Yeah. Yeah. So you, what you could do is try to attract. Even though its quote is blocked by the firewall there, you still probably could attract users from those regions of the world.

And then if you track how divergent their views are, the people who actually live in those places, how divergent the views are, say Taiwan or something. That would be interesting. That would be a kind of interesting little data science exercise to see.

Stephen Grugett: Yeah, it's kind of hard to tell, though. We don't, we don't really have a good estimate of how many people, how many users we have, are in China just because they're all using VPNs from all different [countries].

Steve Hsu: Right.

But I think if you were particularly interested in this question of like divergences between people who are centered in different places, you could ask people, you could say like, hey, where are you? You know, like, what region do you live in? Do you think you have good insight into what's happening in this region?

You could break that out as a separate variable that, you know, you could analyze. not sure what else. I like the fact that these markets exist because it's, basically for the reason I gave earlier. I'm just like, if I'm too lazy to do a really deep dive on something, it's just like, wait, what, what happened?

Oh, and then I just go look at the market price and I see like, how, you know, I just use that as like a first cut on how likely I think this thing is, how real this thing is. Yeah.

Stephen Grugett: It's a, it's a great sanity check.

Steve Hsu: Exactly. So, I like it and it's a, I don't know, maybe, maybe you need a further gamification. I, I don't know, like, It doesn't feel to me, like, I think for some people it's as fun as playing, like, a game, like Bridge or World of Warcraft or something.

But to me it doesn't feel that way. I don't know if that's required, it requires some further gamification to get people to, like, be more interested in playing it. I don't know. That's maybe something to think about.

Stephen Grugett: Yeah, for sure. Yeah.

Steve Hsu: So let's, let's just broaden out since we're, you know, I think we've covered a lot about manifold markets.

So tell us what the scene is. How old are you, Steven, roughly? You don't have to say exactly. I'm in my early thirties. You're in your early thirties. You're living the dream here in San Francisco. You and I were just at a meetup last night with a bunch of interesting people. Founders, machine learning engineers working with robots.

You know, super fun conversation. You're at the heart of this, like the density of interesting people who are trying to do things in the world is no higher than here anywhere else in the world. Just tell us what, what, what it's like to live here.

Stephen Grugett: Yeah, I think you, you captured it a lot. This is the center of the universe for tech and people building the future.

What that means in practice is like at every party you go to, you can expect to talk about AI and your AI timelines, things like that. Yes.

Steve Hsu: Yes. But do you, now do you feel like, for the average, because obviously we're living among, I don't know what the total population of the Bay Area is like, maybe it's 5, 10 million or something, I could be way off on that, I guess.

But, it's millions of people. Do you think like there's the universe and even in Silicon Valley, San Francisco, there's this, like, you know, founder, tech person, rationalist sub world and like how far out do you feel like in terms of, being, unusual compared to the normies?

Stephen Grugett: I see. well first, like Manifold, you know, we, a large chunk of our user base comes from the Bay Area Rationalists. They're kind of their own thing. they're a bit of a separate tribe from, like, SF Tech founders. So we, you know,obviously have substantial overlap with both communities.

The Rationalists are, you know, more, strongly based in Berkeley than here in SF.

Steve Hsu: Yep. But in both places, like, yeah, I guess my impression is, when you're a founder here, like, my first company was, located here, and actually in, in Berkeley, in Emeryville. but this was a long time ago, like, in the first internet bubble.

They're like, for sure, like, if you went to parties or the types of parties that we would get invited to, you'd meet other founders and engineers and venture capitalists and stuff, but still you're like part of like a little sub world. And like you wouldn't expect if you, like some normie person invited you to their party, you might go to that party just cause maybe there's some cute girls there or something.

But, you wouldn't necessarily expect people there to have that much insight into what you were really doing. Like they're just like, Ooh, you're a, you're the CEO of blah, blah, blah. That's awesome. But they wouldn't know much, like they don't, they don't know what a. They don't know what a venture, an A round, what that means and, and things like that.

Stephen Grugett: So, even some of that cultural knowledge, you know, is, you know, has made its way out into everyone in the Bay Area. You know, there are Uber drivers who are asking, you know, about your financing rounds and things like that. That's not, that's pretty common actually.

Steve Hsu: Okay, that's great. Yeah, because in the rest of the country, there is no way.

Like, no way can you go in, you know, in the middle of Chicago and start talking about your, you know, your, your board, the board resolutions we just passed for vesting of options or something. No one would have any idea what you're talking about.

Stephen Grugett: Yeah, yeah, it is one of the unique things and part of the magic of being in the Bay.

Steve Hsu: Yeah, that's great. You know, when we, the most recent funding round, the last couple funding rounds I did with companies that I'm involved in, we were able to do it totally virtually. you know, as partially as a sort of, I think COVID basically moved the whole system into accepting these virtual meetings.

And so we were able to have meetings with investors, like one after the other, just like on, basically on Zoom or Google Meet. But here, of course, you can actually meet everybody face to face, which to me is a novel thing. So I like to come to the Bay Area every now and then just to have these face-to-face meetings.

How about here? Do you, do you now go and take face-to-face meetings when you're, if you were doing a funding round, would you be visiting the, the funds or would you still be doing it over, uh?

Stephen Grugett: Both. I think like the Zoom and, you know, Google Meets are here to stay though. Yeah. It's just, it's so much more convenient for both parties.

Yep. I think like most intro calls are probably gonna be virtual for some time to come. Yep.

Steve Hsu: Yeah, but I think then, then like here it's much easier just like, like see, like just even like the, the short few days I've been here, everything is within, almost like within a walk or a short Uber ride or short Waymo ride, everything else.

And yeah, that's got to be very different

Stephen Grugett: from, yeah. So the fact, yeah, we, we can and have had like physical meetings with our investors and that's, that's great. Yeah,

Steve Hsu: good. Alright. Anything else you want to add about life? What's, what's the weirdest thing that's happened in your Bay Area life recently?

Stephen Grugett: The weirdest thing. There are a lot of weird, unpleasant, weird things. There's been like human excrement in front of my office that I've had to clean up. Oh gosh. Um,this is the flip side of the Bay Area.

Steve Hsu: I love, you know, I always say this to I teach a class on tech entrepreneurship at Michigan State.

I always say this to people. They're a little bit surprised sometimes. I say like the founder has to do, and it's actually good for the founder to demonstrate doing the lowest, worst possible task in the company. It just shows their determination for the company to succeed. But I must say you have exceeded anything I've ever had to do because I never, I never had to clean up human excrement in front of the office.

Stephen Grugett: Yeah, this is being a founder in SF in 2024.

Steve Hsu: I have to say, walking around the city in this week that I've been here, I haven't seen, I have seen homeless people obviously, and also deranged people, and so, but I didn't, maybe I just haven't walked by the monster encampments that I used to.

Stephen Grugett: This is a pretty good part of town. Yeah. Um,but even in the good parts of town, bad things happen occasionally.

Steve Hsu: I see. Yeah. Is it true that they cleaned the city up for Xi Jinping?

Stephen Grugett: I, I believe so. I actually never went over into that part of town during ABEC, but I believe, yeah, I believe that's true.

Steve Hsu: Okay, good. Well, at least he has, he serves some use here in the Bay Area.

Alright, well, Stephen, it's been great having you on the show, and I wish you guys the best. What's your message to someone who's just learning about Manifold Markets? What should they do next?

Stephen Grugett: Give it a try and see how well you can do. You're going to learn something one way or another.

Steve Hsu: Right, you learn something about yourself by playing the game, right?

Stephen Grugett: Exactly.

Steve Hsu: Alright, thanks a lot.

Stephen Grugett: Alright, thanks for having me.