- Aaron Harris, a partner at Y Combinator, interviews Olaf Carlson, the founder of Polychain Capital, about cryptocurrency and blockchain technology.
- Olaf Carlson explains his background in Bitcoin and his experience as the head of risk at Coinbase.
- Polychain Capital is the largest blockchain investing hedge fund in the world, with backing from Union Square Ventures, Dreesen Horowitz, Sequoia, and Founders Fund.
- The fund takes a long-term investing approach and focuses on early-stage investments in cryptocurrencies.
- Olaf Carlson discusses the criteria for evaluating investments, including novel concepts, application of new technologies, and the expertise and ideas of the team behind the project.
- The fund primarily invests in protocol and infrastructure layer projects that enable developers rather than end-user applications.
- Olaf Carlson mentions Filecoin as an example of a project that is needed and adds value to developers in the decentralized web ecosystem.

- Filecoin is an incentive layer for the IPFS server system, which is a content-based addressing network.
- IPFS allows for a decentralized architecture with redundant hosting from nodes across the network.
- Filecoin adds high uptime, fast bandwidth, and resiliency to the IPFS network.
- Web 3 services, like OpenBazaar, can provide a better user experience with the IPFS network and Filecoin.
- Web 3 services have the potential to enable behavior that is not possible on the centralized web.
- DAOs (decentralized autonomous organizations) are examples of web 3 native applications that could be enabled by services like IPFS and Filecoin.
- Cryptocurrencies are often perceived as get-rich-quick schemes, but they are actually novel technologies.
- Understanding blockchain technologies and their complexities is important for grasping the potential of web 3 applications.

- The speaker discusses the need for new tokens in decentralized applications (dApps) and the incentives they provide.
- They argue that while existing highly liquid tokens like Bitcoin could be used for payment in dApps, creating native tokens creates a different set of incentives for developers and early backers.
- The speaker compares tokens to shares in a company, stating that they provide narrow network effects and incentive effects specific to the application.
- They believe that adding monetary incentives to open source projects will accelerate development in the ecosystem.
- The speaker also discusses the question of focused effort versus distributed effort in terms of creativity and the opportunities presented by the coordination of capital through incentives in the cryptocurrency space.
- They note that the current trend of crowd sales and initial coin offerings (ICOs) is still in its early stages but is likely to continue and have a significant impact.

- The speaker discusses the success of Tezos' crowd sale, which raised over $200 million.
- They criticize other projects that have raised large amounts of money without showing much progress.
- The speaker believes that crowd sales should be used to distribute tokens and build a user base, rather than just as a fundraising round.
- They explain the importance of speculation in driving network effects and liquidity in these projects.
- The speaker argues that excess value in crowd sales should go to developers and users, rather than traders.
- They believe that while Tezos may not need $200 million, it is better for the value to go to the developer team than to traders.

- The speaker believes that the value generated from a crowdsale should go to the developer team rather than traders.
- The Tezos crowdsale was uncapped, allowing it to raise $200 million.
- The speaker suggests that a capped crowdsale would prevent short-term speculators from taking advantage of the price difference between the crowdsale and the market price.
- Tezos is expected to go live and be traded on liquid markets in 3 to 4 months.
- The speaker argues that a longer hold period, such as 1-4 years, would incentivize long-term holders and deter financial speculators.
- The speaker mentions being excited about the 0x decentralized exchange and the Maker stablecoin.

- MakerDAO is a decentralized organization acting as a crypto central bank for a decentralized stablecoin.
- The concept of volatility-free crypto coins is fascinating and could be a game-changer for applications like prediction markets and e-commerce.
- Stablecoins, like the one created by the Maker Project, are exciting because they provide a stable store of value.
- Investing in tokens requires thorough research and understanding of the experimental and highly volatile market.
- For those creating tokens, the choice of parent entity and its location globally is a complex and ongoing conversation.
- Resources like the IPFS team and the CoinList project can provide valuable insights for those wanting to dig deeper into the topic.

- Aaron Harris interviews Olaf Carlson, founder of Polychain Capital, about cryptocurrency and blockchain technology.
- Polychain Capital is the largest blockchain investing hedge fund in the world, focusing on early-stage investments in cryptocurrencies.
- The fund evaluates investments based on novel concepts, new technologies, and the expertise of the project team.
- They primarily invest in protocol and infrastructure layer projects that enable developers.
- Olaf Carlson mentions Filecoin as an example of a project that adds value to developers in the decentralized web ecosystem.
- Web 3 services, like OpenBazaar, can provide a better user experience with the IPFS network and Filecoin.
- Decentralized autonomous organizations (DAOs) are examples of web 3 native applications that could be enabled by services like IPFS and Filecoin.
- Cryptocurrencies are novel technologies that require understanding to grasp their potential.
- Creating native tokens in decentralized applications (dApps) provides different incentives for developers and early backers.
- The success of Tezos' crowd sale is discussed, along with the importance of distributing tokens and building a user base.
- The speaker believes that excess value in crowd sales should go to developers and users rather than traders.
- The speaker suggests a capped crowdsale to prevent short-term speculators from taking advantage.
- The speaker mentions being excited about the 0x decentralized exchange and the Maker stablecoin.
- Stablecoins, like the one created by the Maker Project, provide a stable store of value.
- Investing in tokens requires thorough research and understanding of the experimental and volatile market.
- The choice of parent entity and its location globally is a complex conversation for token creators.

Olaf Carlson-Wee is the founder and CEO of Polychain Capital -http://polychain.capital/

Aaron Harris is a Partner at YC - https://twitter.com/harris

YC's podcast is hosted by Craig Cannon - http://twitter.com/craigcannon

I'm a partner here at Y Combinator and I spent a lot of time thinking about fintech and how technology is changing the way we use and move money and how that plays into assets, banking systems, insurance systems, which all kind of comes together for a lot of people in cryptocurrency, which is obviously this new, not super new, but fairly new in the context of money and financial systems set of things.

And I think there's a lot of misunderstanding about it, which is why I'm happy we're doing a bunch of podcasts, talking to people who really get this stuff, because I have a ton to learn.

So, Olaf, I'm really super excited to talk to you about this, because you get it way better than I do.

I'm the founder of Polychain Capital, which is the largest blockchain investing hedge fund in the world.

Far before Polychain Capital, I found out about Bitcoin in the summer of 2011 and became extremely infatuated with this concept of digital assets.

I decided to pen my undergraduate thesis on Bitcoin that year.

I think this was probably 1 of the first academic works on cryptocurrency.

After that I joined Coinbase as the first employee.

Coinbase was actually a Y Combinator company and is now raising money purportedly at over a $1 billion valuation.

I was at Coinbase for 3 and a half years, and during that time I was the head of risk.

Day 1 for the entirety of my time there, I was paid exclusively in Bitcoin.

So there were good times when that was a good decision and times when that was a bad decision, but I think nat-nat it was right.

And I left last summer to launch the Polychain fund.

So we launched with a little, around 4 million under management, and now we have about 200 million under management.

We have backing from Union Square Ventures and Dreesen Horowitz, Sequoia and Founders Fund, among others.

So, a somewhat unusual LP base, and we can get into that a little bit.

Are those funds functioning as straight LPs or are they investing in the GP?

So, I worked for a hedge fund for a while, so I have a conception of how that looks in my head, the kind of things that you look at, but there are a lot of different kinds of hedge funds.

There are macro hedge funds, there are long short funds, there are activist funds.

What does it mean to be a hedge fund focused on the blockchain?

Yeah, so a lot of those strategies you just stated could all apply to cryptocurrency assets.

Some of them are mechanically much more difficult.

So for example, short positions in the space are difficult to achieve in a secure way.

Also, for example, something as simple as margin trading can be hard to find for the right assets that you're trying to trade in.

For us, We're sort of what I would call fundamental investors in that we're long only with a long holding period.

And I spend most of my time reading research papers about cryptocurrencies.

So it really depends on how fast things grow.

So we are rebalancing the portfolio roughly every 90 days.

So the goal is to hold things that are good positions for years.

But we may rebalance or sell down on those positions over time.

So in terms of pace of movement, it kind of looks less like a hedge fund and more like I think how people would think about almost a mutual fund kind of thing.

But I'm guessing hedge fund because it allows you a lot more flexibility in terms of what you're allowed to invest in and the strategies you can pursue?

So the tricky thing was I knew in my head what I wanted to do as far as investing in these assets.

And that it actually looks a little bit like a venture investment in that you find things extremely early when they're still just a specification.

So these protocol specifications or white papers are extremely detailed descriptions of exactly what a protocol will do.

It's often very possible to understand a lot about what this will look like just from that specification.

And I often feel comfortable investing if I meet the team and I read the specification and really like what they're doing.

So that looks very much like a venture investment in that you're investing more on an idea than you are on proven data or numbers.

But from there, these investments very quickly graduate into liquid markets.

And so because these things are liquidly traded on an order book, unlike a venture firm, you really have to balance those positions.

You can't just sort of wait for the IPO, so to speak.

So because of that, I knew we needed to manage liquidity, and I knew we needed to have really good risk management around our positions because we have month-to-month volatility, unlike a private equity or venture fund, which can just kind of sit on positions forever.

But I wanted to take this long-term investing approach that a venture fund would, where a lot of your success is defined by very, very big winners and positions that rise substantially in value.

So we invest in things at a really early stage.

We form really good relationships with entrepreneurs.

I'm really proud of a lot of the people we've backed.

And then, you know, hope to hold those positions for years as those things grow.

So can you dig into the technical aspect of what you're actually looking for when you're reading the white papers?

So when we're reading these protocol specifications or white papers, we're looking for basically a novel concept or novel idea.

So this could be a new application of existing technology.

This could be an application of technology that hadn't been tried in a cryptocurrency framework before.

This could be just a kind of new game theoretic model that no 1 had ever designed before that allows for a novel incentive system or incentive structure, new consensus systems, new just kind of like anything in this category of enabling new behaviors.

So there's a lot of technologies out there that are forks of existing technology.

So they'll fork Bitcoin into something very, very similar.

That's usually not very interesting for us.

We really, we really like things that are sort of written from the ground up.

So like Ethereum was not a fork of Bitcoin.

Tezos, whose crowd sale actually ends today, is also written from the ground up.

It is not a fork of Ethereum, even though it has a Turing complete scripting language like Ethereum.

so when you're vetting a team, given that there are very few, you know, cryptocurrencies with a long track record, are you looking for like how many times they've committed to evaluate them or like what are you looking for

yeah we we definitely look at the code base I I think a lot of it though you know just pragmatically speaking like we we can't review every line of code in a new blockchain, right?

That's, we do require all our investments to get security audits from, you know, outside third parties, but even that, right?

These are often experimental technologies and they may be broken in unforeseen ways.

So there are what I would call unknown unknowns here.

There are things that you kind of don't know that you don't know.

That said, I think that I have a long history of kind of, like at Coinbase, I did probably about 500 interviews.

I'm very used to talking with technical people and getting a sense of their subject matter expertise.

I also really feel comfortable betting on people if they have a great idea.

So I think that it's like seed investing.

A lot of the time you meet a really great entrepreneur, they have a really great idea, and that's basically what you're investing in.

Scottie Pants So there's a dividing line or a difference between the technology being interesting and being fundamentally new in some way, and there being a use for that technology down the road.

And sometimes the uses are completely opaque, because we don't know what the technology is going to do to the world, right?

The things that you invest in that do best actually change the world and create the conditions for which they will be successful.

How do you evaluate that set of things, or how do you think about that set of things?

Yeah, So the interesting thing here is that most of what we're investing in today are protocol or infrastructure layer things that really aren't meant to be end user applications.

Most of what we're investing in are really almost like developer tools to be used for end user applications.

So a lot of what we're thinking about is what's going to enable developers more than what is going to be interesting for users.

Scott Winslow But in that case, the developers are your users.

So I think it comes back to the same question of, do you have a view of what the world looks like in 5 or 10 years?

The things that developers will need to be able to do that guide your thoughts on, okay, this is not only an interesting technology, this is a cool thing, but it will allow this piece of the future that I believe in.

Yeah, so it's so early in the development of what I would call Web 3, or this kind of user-owned decentralized web, that oftentimes right now to me, it's pretty clear which things are really needed and which components of this kind of infrastructure layer are obvious value add to developers.

So you know, For example, a project like Filecoin, which is launching pretty soon, created by Juan Benet, who actually was also a YCE alum.

Filecoin is basically decentralized server client architecture.

So a Filecoin is an incentive layer for the IPFS server system.

So IPFS, instead of being location-based addressing, like IP addresses, it's actually content-based addressing.

So when you click a link in the IPFS network, instead of being routed to a specific server, you're routed to the nearest server that has a specific piece of content.

What that means is that you can have a totally decentralized architecture where you have redundant hosting from many nodes across the network and you pay Filecoin in order to submit requests to those nodes.

So it's like a distributed server architecture.

Right now, you see a lot of distributed services, say, OpenBazaar.

OpenBazaar is a peer-to-peer marketplace, sort of like a peer-to-peer eBay.

Well, OpenBazaar right now, as a user, you actually have to download an application like a DMG to your computer and run it locally and basically run a node in the network.

This means that in web 3, you'd have to download a new application sort of like it is on mobile instead of being able to just go to a URL in the browser.

With the IPFS network in place and Filecoin which basically adds high uptime and fast bandwidth and resiliency to the IPFS network because it has this incentive layer.

IPFS is mostly hobbyists right now before that Filecoin incentive layer gets added.

Then when you visit OpenBazaar in the browser, you can have a better user experience and it's like a more clean experience of this Web3.

So Filecoin to me becomes a sort of clear infrastructural component that's just a huge missing layer here right now.

So does that mean that the basically there's like the use case for IPFS and Filecoin is predicated on OpenBazaar being a thing or something like OpenBazaar, there will be other things.

Is then Is the argument then, okay, an open bazaar is fundamentally a better solution to the problem of buying and selling things than today's centralized systems like eBay?

Do you have to believe that to be true, or is there some other world in which, no, these things actually exist side by side, some people use 1, some people use the other, and because the economy is just so damn big and this thing is just, it's all just gonna get bigger, there's 2 winners.

Yeah, so I think that over the long term it's possible a lot of Web3 services will compete on cost structures with centralized services like say eBay.

But that's actually less interesting to me.

What's interesting is what these kind of web 3 services do that a normal web service could not do.

Because the eBay experience is actually pretty solid for buyers and sellers.

I think the fees they charge are reasonable given they have like services, they, you know, value add services like dispute arbitration.

Because all those sorts of things are kind of then independent on something like the OpenBazaar platform.

But I do think there are going to be web 3 apps that really are uniquely enabling behavior.

So types of things that we really can't get out of Web 2.0 or like the centralized web.

But I do have this intuitive feeling that we actually don't really know what a lot of those use cases are.

And what we're talking about is services for things like DAOs, a DAO here being a decentralized autonomous organization That's basically a pooling of capital on the blockchain When you think about a DAO pool of capital, it can't I don't think it can realistically engage in traditional legal arrangements.

But it can arrange, you know, enter into smart contract, or like software based legal arrangements.

So I actually think that you'll see more and more kind of services for DAOs with DAOs here being just basically global pools of capital that exist only in the blockchain and not an illegal entity.

They sound a little bit sci-fi right now, but I think in 5 years we could see some massive, massive kind of, you know, it's like in the early internet, I think it was hard to imagine Facebook, right?

And I think there's this video I've seen of Mark Andreessen explaining Netscape in like 1994, and an audience member says, what are some websites you think are interesting?

He really stumbles, like he really struggles to answer the question.

And he kind of comes back to, well, this is really great technology.

Today I don't really know what to point to.

And I feel like pointing to things that are ostensibly better versions of the centralized web or efficiency gains on the centralized web is sort of the easy path.

It's kind of like, oh, let's take something like a library and put it on the internet.

The kind of layer 2 or like native internet application is Wikipedia, right?

Which you couldn't really have in a non-internet environment.

So I think we're going to, at the beginning, be comparing or trying to port rather web to things onto this web 3 or decentralized user owned web but that over time we'll find those web 3 native applications.

And those are the things that I really care about.

Even though right now they do sound somewhat sci-fi, and I think it's very unclear what they look like.

Yeah, I think there's kind of this, the way that I've tried to think about it, and I'd love to know if this is wrong, because it probably is, is that the argument being made now is sort of that there's this protean mass of things happening.

It's like the primordial goop from which life originated.

The blockchain technologies being built now are the early amino acids.

And it's entirely possible the whole thing is going to go to 0, right?

That nothing will happen out of this attempt.

But if you own a piece of the right amino acids, essentially those are going to create something.

Something is going to happen when you have enough creative energy focused on a small enough surface area that we will see.

But yeah, but even like in addition to like some people making money, like I think it makes sense now to talk about what we were talking about before we started, which is like, what are people not understanding in like what's covered it?

How is how's cryptocurrency portrayed in the media and then what are things that people are essentially getting wrong or not covering that are hurting people from understanding like the creative power that they might be able to have.

Yeah so this this makes me think of something that I saw Naval Ravikant, who's a backer of mine, tweet.

And he's a great thinker in the kind of cryptocurrency space.

And he said something along the lines of, you know, Bitcoin is, you know, the largest fear of kings and those in power across the world dressed up as a get rich quick scheme.

And I really do think this idea that in the mainstream kind of cultural consciousness, cryptocurrencies I think mostly are perceived as a get-rich-quick scheme.

And I actually think it's kind of an amazing bootstrapping mechanism to get a lot of speculation into the space, liquid markets, incentives for developers, and ultimately people building on these technologies.

So I'm actually not necessarily against per se a lot of the speculation happening.

However, I do think the mainstream media mostly views these as financial markets that are novel rather than novel technologies, if that makes sense.

What are the key understandings that people don't get?

Yeah, so I mean, I think that these technologies are complicated.

So I think a lot of people were reeling to understand Bitcoin and just the premise of a blockchain and like right when they thought they maybe Understood it all of a sudden everyone's talking about aetherium, and now they're trying to wrap their head around a Turing complete Language which can actually execute arbitrarily complex software in the blockchain and all of a sudden this concept of money or currency is a very limited metaphor for what these things are doing.

And kind of, you know, what we're going to see is right when you're kind of starting to understand Ethereum and ERC-20 tokens, which are digital assets built on top of Ethereum, that don't have their own blockchain but are kind of secured by the Ethereum network.

And it all is kind of an order of magnitude more complex than Bitcoin.

We're now going to see just even more complicated things happening.

We're going to see Tezos, which has on-chain governance where it's an inward-looking protocol that can actually use protocol rules to change itself And something like Tezos can actually build pools of capital And incentivize developers to build on the protocol and then sort of dilute every holder of the Tezos tokens in order to reward those developers.

And so now you have a protocol that's incentivizes development of itself.

And this is all like on a native protocol layer, A lot of these things, you know, that's like an amazing recursive feedback loop where the protocol actually builds itself in a strange way.

But a lot of these things are going to be again, another order of magnitude, harder to wrap your head around like things like DAOs, which I actually think are going to grow a lot over the next year or 2.

So I think that the kind of technology is always going to be basically a step ahead of the cultural understanding of the technology.

I don't expect everyone to spend their days reading these protocol specifications and trying to really wrap their head around it.

But I think that if you don't dive into the technology and how it works, you can just ultimately never really have a deep understanding of this.

So this kind of is related to 1 of the questions someone sent in for you on Twitter.

And he writes, most, maybe all, AppCoin tokens are simply used to pay for some resource, like compute, hard drive space, energy, etc.

Why is a new token needed in these cases and why can't we already use existing highly liquid tokens like Bitcoin to pay for these?

so I think it's actually a great question from John who I actually know in real life.

So I think that the disagreement that I would have is that a lot of people view tokens from a pure technology perspective.

So they think, we can do this with Bitcoin and like the Lightning Network on Bitcoin or something like that, which has a high throughput layer 2 network.

And I think there is no strong reason from a technology standpoint.

You could use Bitcoin for a lot of these different services or rather Ether for a lot of these different services.

However when you create a native token you create a very different set of incentives.

So for the developers who are building that network, you create a real narrow incentive around that application or around that token, right?

So that it's not just tied to the larger value of Ethereum, which if they kind of contribute this great application on top of Ethereum, Ether might grow in value by 5 or 10%, but that's not really a great outcome for that founder or developer.

In addition, the early backers of that application or token that's part of this app coin as as John put it aren't if they're just You know contributing to this product or service with ether They really don't have this embedded incentive effect for that network to grow so tokens in this sense are really more of an incentive structuring or like a kind of game theory hack, right, to get really powerful network effects around a specific application.

But I sensibly, you know, in the United States, we all sort of benefit from the strength of the US dollar.

But when I create a new company, right, I create shares specific to that company.

Because although even though I might help the US economy and thus help the US dollar or kind of the underlying network as you might think about that, I want narrow network effects around what I'm building.

And so I want my early backers or investors to have upside relative to my specific application, not just to the strength of the US dollar.

So it's a little bit of a rough metaphor.

I don't think it's perfect, But I do think that tokens provide really narrow network effects and narrow incentive effects, and that's what's so important about them.

It's not really a technological reason, but rather a sort of game theoretic reason.

So There are 2 arguments baked into that that I'm not sure how to think about.

1 of which is the incentive structure question itself.

A lot of the underlying message here is that money is, at the end of the day, the best incentive that you can give people to develop something.

And I think there are very different opinions on this question and I know people who are motivated by money and people who aren't and some of those people on either side have achieved great things in life both for themselves and for the human race.

So that's sort of the first thing, is this the right path to say, hey, we're gonna break everything down, everything that, all this fundamental technology should be motivated around money?

The second question that I have is around this idea of empowering the collective, right, to develop new things, and sort of this massed creativity versus focused creativity.

And I don't know that these are in direct opposition to 1 another, But 1 of the things you get with a central authority that does something and says, hey, you're worth this much, you're worth that much.

There's always going to be problems with the distribution of incentive there.

But the thing that you get on the other side is focus of intent, which sometimes is bad when it heads in the wrong direction and sometimes is good when it's in sort of a positive direction.

I think the Manhattan Project is like directed focus on like, hey, here's the thing that we all need to do, all you great scientists.

Yes, there are subgroups within that, but here's what we're going to do.

So I'm curious how you think of both of those things, like, in terms of how they incentivize positive growth, I guess, and if those are the right ways to think about it, or if there's another way that I should look at this?

So I think, you know, I think we're seeing for the first time capitalism and like money incentives being built into open source projects, which I don't think we've ever seen before.

I always cringe a little bit inside when I see the Wikipedia banner that's sort of like, please donate $2.

And I can't help but think there's a world where Wikipedia was like incentivized with a token and Jimmy Wales for his contribution to the world actually captured a huge amount of value.

In that he created a lot of value, but he did not capture a lot of value, obviously, like having to go out hat in hand for $2 donations.

Whereas I think that what we're seeing now with open source projects is the creators of something like Ethereum, they're creating a huge amount of value for the world and they're also capturing that value.

And I think that value capture and that kind of adding this kind of capitalist for-profit mentality to open source will absolutely accelerate development in this ecosystem.

I just think I don't have any question in my mind that monetary incentives drive growth and drive people to work on a project.

You know, everyone, you know, or most people in the United States like have a job and they mostly do that for the money at the end of the Day, that's like a big driver or incentive No matter how much someone is passionate about something if you pulled out all the money from it, it would be very hard for them to continue doing that in most cases.

So I think adding monetary incentives to this, while it does draw in a sort of different group of people in some ways than you'd find in traditional open source communities like hacking on the Linux kernel or something.

I do think that it ultimately will dramatically increase the pace of development here.

See all Y Combinator transcripts on Youtube

Blockchain Investing - Olaf Carlson-Wee and Aaron Harris