Video - Bitcoin Fireside Chat with Marc Andreessen and Balaji Srinivasan

Marc Andreessen and Balaji Srinivasan talks about the dawn of Bitcoin and where it begun. They also discussed how it has been second to burst since the internet.



Sheryl Sandberg:  Hello, everyone.  So we're having a fireside chat without the fire.  And I want to start off by talking about there's always exciting things happening in Bitcoin, whether it's Mt. Gox imploding or people being arrested or getting subpoenas from the New York financial regulator.  And I was just curious as investors in this space whether there's more or different due diligence you have to do when you're looking at the companies that you might want to put money into.

Marc Andreessen:  Yeah.  So I'll just start by saying two things.  So maybe they can turn on the echo just a little bit if that would be possible.  I'll try to talk against myself.  So, kind of the two things that we think about a lot, so first of all, we think it's important put Bitcoins sort of in context with where it actually is.  And so Bitcoin is, right, Bitcoins only existed for five years.  The original paper came out in 2009.  I think the relevant comparison point for Bitcoin is actually about 1993 for the consumer internet and -- or 1994, actually.  And I say that because like 1989 was actually when the web was actually first invented when HTTP and HTML were invented.  And there was like this five-year period between '89 and '94 where those of us who were working and kind of on the early, you know, at the time in a SAFE Net and an early Mosaic.  You know, you could kind of use this stuff and you could see it, but it was really fringe and really new, and really weird, and really scary, and really odd.  And then it kind of arrived as a fringe technology.

And then by the way it arrived with fringe politics and it arrived with fringe characters and there were actually a whole cast of characters in that era.  Some of whom became actually very successful, some of whom are historical figures now.  But there were a whole set of sort of fringe characters who had these sort of crazy ideas.  Many of the actual technical ideas turned out to be absolutely correct.  Many of the political ideas are depending on your point of view, they're not to be correct or not correct.  And actually, a fair number of people got disillusioned politically kind of through that process.  But it worked and there was this process of maturation through which it worked.  And so I think that the critique of Bitcoin as sort of being fringe technology accompanied by fringe politics and fringe characters, I don't know how you get fringe technology without fringe politics and fringe characters.  You just have to go through a maturation process where you come out the other end when the fringe technology goes mainstream and gets widely adopted.  Along the way, the fringe characters and the fringe politics tend to get alienated and then tend to move on to the next fringe technology, and the cycle repeats.  But you don't get the new technologies from the mainstream, you get them from the fringe.

And so I think this is a case study of that every single one of the kinds of things that you mentioned is an example of kind of how I think the early stages of this adoption cycle happen.  So it would be really shocking if this kind of thing weren't happening.  It would be very historically unusual if this kind of thing weren't happening.  Speaking for us, we consider ourselves mainstream investors and we consider ourselves in the business of building mainstream companies.  We only invest in companies that want to be on the regulatory straight and narrow.  And so the thing like we don't take any chances on for something that has like implications like Bitcoin is we would not take chances on somebody who's going to run an exchange the way that Mt. Gox got run or anything like that.  Like, you know, Coinbases are big public investment and they're extremely rigorous about being properly licensed.  And so everything we do will kind of fall in that category.

Balaji Srinivasan:  Yeah.  None of these guys past our diligence.

Sheryl Sandberg:  And Coinbase was your first big Bitcoin investment.  It's interesting that the signal to noise on Bitcoins.  It's just one of the many companies that you're investing and we're talking about Oculus Rift out there, but you talk so much more about it.  Does it has to come to really just dominate your talking points in a way that you didn't expect or did you always know it would be like this?

Balaji Srinivasan:  Well, Bitcoin, I mean, is a phenomenon, right.  It's not just -- I mean, Oculus will also become an Echo-system and so on.  But Bitcoin is something where we think of it as potentially on the scale of the next internet.  So as such necessarily is going to occupy more and more of our time and a lot of our portfolio companies integrated.  So it's not like just one technology in one company.  It's kind of cross-cutting.  So that's kind of -- we think of Bitcoin like mobile.  It's not one company, it's something that is broad.

Sheryl Sandberg:  And how did you decide on Coinbase as the beginning?

Balaji Srinivasan:  So they've kind of gone through a national selection process, right.  Like, in the sense that we've been tracking them for a while.  And they're very smart guys who had implemented a ton of features on top of both enterprise features like the Merchant A.P.I. and so on, as well as consumer features like the ability to purchase Bitcoin on the website while maintaining regulatory compliance on a shoestring budget, right.  Anybody who can do a lot on a little is somebody who is a good candidate for venture investment because if they can do this much on, say, 1 million or 500,000, what can they do on?  Much more than that, right.  So it's a kind of a national selection process.

Sheryl Sandberg:  And then I assume you guys were not done investing in Bitcoin companies?

Marc Andreessen:  Yeah.

Sheryl Sandberg:  What are you looking to put money into?

Marc Andreessen:  Yes.  So I'll give a general answer.

Balaji Srinivasan:  Yeah.  Yeah, sure.

Marc Andreessen:  And Balaji has specific ideas.  So the general answer is we think there's a whole Echo-system here.  And, I mean, we think the Echo-system is already forming.  I mean, in a lot of ways we view ourselves as reacting which is basically we're now seeing hundreds of high-quality entrepreneurs and high-quality technologists, you know, kind of people you consider to be top tier people in any field who are working on all kinds of ideas.  And, you know, sort of ideas spread the full gamut or everything you can imagine.  And so we do think of it as an Echo-system.  Again, I'll go back to the internet analogy.  In 1994, as a venture capital firm, it would have been a good idea to take the internet seriously and it would have been a good idea to invest in a cross-section of internet infrastructure companies, internet application companies, internet content companies.  You could have made a whole series of batch.  The venture firms did do that then, you know, extremely well.  And so our goal here is to do that, basically build a portfolio, diversified batch in all the different areas of the echo-system.

Balaji Srinivasan:  Yeah.  And I think on certain things we're looking at, specifically, one of the things we've thought a lot about is that Bitcoin is like an inverted pyramid right now.  There's a lot of news coverage nowadays, right.  And there's an increasing number of companies and investors.  There is relatively a few people who are working on the code protocol and that kind of activity, right, like, not that many developers who are actually working on  And so, there's two kinds of vehicles that we're looking for in particular to fix that.  People in the audience might be interested in doing things like this.

So one is like a Red Hat for Bitcoin, right.  So Red Hat was seminal company in the early days of the internet.  You Know, founded in 1993.  They professionalized Linux development, both creating both user friendly Linux situations as well as later Red Hat Enterprise Linux and enterprise-friendly distributions.  So a vehicle like that, a Red Hat for Bitcoin, I think, would be a very interesting thing to fund and to look at something that had a critical mass of code developers.  And while in theory, a consulting company may not make that much money in practice Red Hat has made like.  It's $11-billion-dollar market cap.  And I think these very few things are going to grow like Bitcoin consulting over the next three to five years.  Everybody's going to need a wallet on their website to accept payments.  It's going to be something where millions of websites are going to be retrofitted to accept this.  We're going to get into that sort of things.  You can get it all over the place, so Bitcoin consultant's growth area.  Moreover, there's all kinds of new ways to monetize 20 years later on the internet.  You could have a Red Hat for Bitcoin that, say, had ads or affiliate codes on the download site for the Bitcoin wallets.  You could do things like Teespring to monetize and many different ways to monetize.  And Bitcoin itself gives ways to monetize new monetization models for open source.

And the second kind of company that we'd like to fund in addition to Red Hat for Bitcoin is like an Underwriters Laboratories for Bitcoin.  So under those laboratories, it does kind of electronics certification, like, safety testing and so on.  So development of something along the lines of PCI compliance standards for Bitcoin as well as the inspections to check that security and stuff like that is well done on these kind of sites.  I think it's going to be a very important company that's going to come out there.  And the thing about that is we've already had that circle organically happening part of that, like, recently Coinbase and Blockchain did a home and home or each of them went and looked at each other's security and then publish the results and so on.  And I think an organization like that that sets up the practice is going to be pretty important.  So these are the two things in particular that we're thinking about.

Sheryl Sandberg:  Okay.  The Ernst & Young for Bitcoin.

Balaji Srinivasan:  Well, yeah, sort of.  I prefer Underwriters Labs because it's a little more technical.  But, yes, some are audits.

Sheryl Sandberg:  Anything else that you think is missing from the Bitcoin echo-system?

Balaji Srinivasan:  Seeing a ton of companies out there.  I mean, there's a lot of very interesting things out there, but I do think that people should focus on the core infrastructure and things that are little bit lower level or maybe more interesting to us because that's underinvested in or thought about right now.

Sheryl Sandberg:  Okay.

Marc Andreessen:  Yeah, the only thing I'd say it's like we're getting increasingly amped up over the idea.  Basically, Bitcoin is an enabling technology for other technologies.  And especially for machine to machine kinds of things.

Balaji Srinivasan:  Uh-huh.

Marc Andreessen:  And so there's all these scenarios where machines talk to other machines and there's all these issues with resource contention and who gets priority on a phone call or a video stream or who gets to fight you.  But my favorite example is your car is driving on the street, your car now has a wireless connection and an operating system in it.  And your smartphone, which has a wireless connection operating system, and you've got driving -- driving an extra parking garage and, like, do you have number one, any idea of whether there's a spot in the parking garage?  It's open, which is sort of question number one.  And then question number two is, boy, I'd sure like to be able to just buy that.  And when I pull in like pull in and park the car, get out of the car, leave, and have the transaction happen automatically.  And in a place like San Francisco where there's always more cars than there are parking lots, maybe there should be a real-time auction happening between cars and garages for allocation of parking spots.

Today, you'd have no technical way whatsoever implement that it's not like you're going to do that with credit cards.  It's not feasible for a variety of reasons.  But with Bitcoin, it would be completely feasible to do a real-time auction between the car and the garage.  And just kind of extrapolate that example across thousands, ultimately millions of cases, where you've got sort of limited resources, you've got potentially unlimited demand, you've got the ability to use prices to basically set a marketplace.  And since Bitcoin can scale way down in value, you could do this with very, very small amounts of currency.

Sheryl Sandberg:  Why?

Marc Andreessen:  I mean, the anti-spam idea is the other one that we always love.

Sheryl Sandberg:  Right.

Marc Andreessen:  It may be too late to fix email, but like clearly the way to fix email spam the entire time was to have there be like a micro charge for every message sent, where as normal people sending a hundred emails a day or a thousand emails a day would make no difference, it would be running or it would be fraction of a penny.  But for spammers to send billions of emails, it would bankrupt them.  And this idea is an old idea.  And 20 years ago, we were talking about this idea.  We just had no way to implement the payments behind it in order to do that.  And basically, every new social system that comes online now has a spam issue and so I do think this is the kind of enabling technology that would make whole new kinds of anti-spam approaches work.  And so just to extrapolate those ideas broadly across anything where you have limited resources and you want to sort of use markets to make things work.

Sheryl Sandberg:  I think the micropayment feature of Bitcoin and the small fractions makes a lot of sense with cutting down on spam.  Why do you think Bitcoin is being central to this kind of parking garage auction?

Balaji Srinivasan:  Yeah.

Marc Andreessen:  Yeah, go ahead.

Balaji Srinivasan:  Yeah.  So -- well, the thing is that for the first time you can have a wallet file that is -- I mean, you can hit keys and set up a program that contains money.  You don't need or let's say a monetary -- something of monetary value.  You don't need a bank to allow you to do that and so on.  You can transmit across a network.  So it decentralizes the ability to write programs that involve monetary value.  And as an analogy, one of the reasons the internet was so exciting is before the internet you had the central backbone operator, the AT&T or what have you, that you had to do a business development deal with in order to deploy code on the network that dealt with communications.  And afterwards you could write programs that took one node and not only sent to other nodes but programmatically exchange packets back and forth with things like P2P, online forums, and so on.  We're not really possible when you're working with AT&T.

In kind of the same way, we have this sort of legacy financial system where five, six years ago in order to send to me a monetary value you had to work with Wells Fargo, you had to work with Visa etcetera, etcetera.  But now you can write a program that doesn't just work with one person sending value from one person to another.  It can be a P2P kind of thing.  You can involve machines, you can involve nonhumans, and you don't need the approval of a bank to do that.  You can have a hobbyist doing it in their garage.  So I think the barrier to entry to writing these kinds of programs in theory.  You could do it in theory, you could always use your PayPal API and so on.  In practice, it's going to be much easier to do with Bitcoin.  That's one.  Then number two is the fact that you can subdivide payments down to Satoshi to, like, 100 million.  And that the payments are nonreversible means that a lot of machine to machine applications now become open that weren't before, right.  So those are kind of three reasons.

Sheryl Sandberg:  So Bitcoin is this incredible transformational technology, yet it was around for about four years before the mainstream business community started investing in it, becoming interested in it.  And the illicit business community seemed to recognize the promise of Bitcoin much earlier.  Why did it take so long for guys like you to come around?

Balaji Srinivasan:  I think that most technology is, you know, start out like the Wright Brothers with their aircraft.  There's not really much to look at at the beginning, right.  Like, early cars had explosions and the early internet certainly had a lot of interesting characters on it for quite a while before normal people got onto it by the mid and late '90s.  I think there's a few reasons in particular for Bitcoin.  One is that the founder remains pseudonymous.  And most of the time, where a company or any kind of open source project whether it's Linus with Linux or Larry Page with Google, the founders out there and they're putting their name on the line and they've got some credibility and so on it builds.  But so pseudonymous founder meant the code had to speak for itself and the theory had to feed for itself.

The second aspect is for a protocol like this, something that is sending monetary value.  The theory should work, but in practice it should also work.  And so having like four years where people were bashing on it in different ways I think it's very important to seeing that it actually work in the real world.  People had a strong incentive to try various kinds of exploit.  So far they haven't really gone too far.  And that's a major sign of strength.  And finally I think the 2011 crash, where the price of Bitcoin rose to 33 and then came down to two and then came back up was an external indicators of the strength of the community of Bitcoin, right.  It's not going to let it die.  And so that weefio moment if you've seen our recent blog post read like, we kind of knew there's a resilience to the community there.  So those are kind of reasons why I think it is the right time to get involved in a big way last year.

Marc Andreessen:  I also say I think the illicit use stuff is overblown.  So one of the funniest things in the last six months has been this senator in West Virginia, Joe Manchin, decided to distinguish himself by writing a letter demanding the immediate banning a Bitcoin because it's used for all these illicit purposes.  Jared Polis who's a young congressman who's extremely sharp, wrote basically the exact same letter to the exact same -- to the Treasury Department, basically calling for an immediate ban on the U.S. dollar.  Using the exact same arguments, right, which is you have this currency, of this anonymous currency, the U.S. $100-bill that can be freely transported all over the world without anybody's approval and can be used to buy drugs and guns and, in fact, today is used to buy drugs and guns at a million times the rate of any sort of online payment mechanism.  The U.S. $100-bill is the enabling technology for worldwide crime and terrorism.  And I think, frankly, he had a pretty good point.  And so I think it illustrates kind of this -- kind of I use the term weird and scary.  The new things just look weird and scary.  And then there's one idiot apparently in Noe Valley as it turns out with an online drug thing or something, and then it becomes this enormously attractive story.  But the illicit stuff turns out it's very sexy story at the time.  It turns out to not be very big part of it.  The exact same thing happened in the early '90s with the internet.  There was -- if you read the coverage from sort of 1991s and 1994, it was story after story after story of child porn and this and you could buy drugs and it was like -- and it was just like this horror show.  For God's sake, you wouldn't want to let your teenager going this crazy internet thing because like who knows what would happen.

I mean, it was just an endless litany of all the social ills that we're going to come out of it as people were getting used to the idea of something new.  And then by 1995, 1996, everybody went, "Oh, this is actually pretty cool."  And so I think we're somewhere in that, but I think the illicit stuffs overblown.  I would also say I think the illicit stuff, our conversations with people in the government who do work involving financing of crime and terrorism.  And we've got some very interesting conversations.  The smart people in the government have figured out that Bitcoin may be an enormous ally to them in particular for a counterterror, for their counterterror programs because the idea of having a Public Ledger that has pseudonyms that all transactions pass through and can be inspected is tremendously attractive to people who specialize, for example, in large scale network analysis who may or may not have been in the press recently.  And again, as contrasted to people using paper currency or people using gold or people using cocaine or people using other kinds of effectively underground currencies, like the U.S. dollar, it's actually a fairly attractive thing to be able to mine the Blockchain, to be able to uncover those things.  My prediction actually is libertarians are actually going to turn on Bitcoin.  I think that's like two years out and it'll be part of the mainstreaming.  The Libertarians will basically discover that the Blockchain is public and that will cause just like enormous freak out.

Sheryl Sandberg:  Well then they'll move on to Zerocoin.

Marc Andreessen:  Glenn Beck is going to get very upset.  Right.  Right. Yeah.

Sheryl Sandberg:  Well, that you mentioned not knowing who the founder is.  Recently, Newsweek said that they thought that they had found the founder.  It's Dorian Nakamoto, a man in California.  And you guys put the article on Rap Genius and kind of savaged it.  I was wondering if you guys think you know who Nakamoto is and whether you think it matters who it is to the investing community to the businesses that are in Bitcoin?

Balaji Srinivasan:  I'd say, Sheryl, I don't think it matters right now.  I think if Satoshi ever came forward, certainly, we'd fund him, right, or her or them.  But module of that, I think it's kind of -- in statistics you've got this conserve base in probability.  And once you have enough measurements from something, you've shifted over and you don't need the prior distribution anymore.  And so the code has taken enough looks at it and it's been deployed in enough places and it's support enough transactions and so on and so forth that the author of it is not the way you judge it anymore.  I mean, if you think about it, you go and read a math book and you look at a formula and rest up to look at or right, like you can see there it lies, right, like you can look at the formula and check it for itself.  You don't need to go and say, "Oh, Newton figured this one out, Alejandra figure this one out."  It's not argument from authority, it's argument from cryptography, right.

So that's kind of reason number one, but I don't think the founder matters in terms of who the founder actually is or everybody has kinds of theories about this.  I personally think that if you read all of Satoshi's emails and in a public forum postings and stuff, these remarks he makes like or she or them where they say for example degrees of freedom when they mean degrees of separation, right.  That's an indicator of often a physicist and that kind of fits with a model of say somebody who is on Wall Street and who is a quant.  And a lot of quants have physics background, is pretty sophisticated about finance who could see plus-plus because they've got a lot of high frequency training kinds of things, who also knew cryptography.  So if you take a quant of Wall Street and who knows cryptography.  That sort of most likely suspect I think.  And it's completely different from Dorian Nakamoto who is the most ludicrous candidate, I mean, somebody who goes to such extents to retain anonymity who signs all their transactions with their middle name.  It completely blows off my mind to think that that was put forward as a reasonable thing that people actually believed it.  So that's why we critiqued it.

Marc Andreessen:  Yeah.  I would also say I've been talking a lot recently about there was a British author -- chemist actually, chemist and novelist named C.P. Snow, who wrote in the middle of the 20th Century, sort of at the height when physics was kind of the new technology everybody was freaked out about, so the bomb and nuclear energy and all these things.  And he sort of wrote this famous essay, we talked about the two cultures.  And at the time what he was talking about was two cultures, a sort of science versus the literary culture.  And basically, he described this kind of cultural divide that was forming where basically -- and it basically mirrors exactly what happened -- with what's happening today.  We're basically the sort of engineering culture, science culture, technical culture, math culture, kind of feels like it's running away with the future and sort of feels very confident and sort of feels like it's discovered the secret formula to progress and then sort of the literary culture, liberal arts culture, non-technical culture of sort of normal people who aren't engineers increasingly feels like this technical stuff is getting like really weird and scary.  And, I mean, the atomic bomb Bitcoin like interesting analogy.

But these sort of fundamental technology breakthroughs look like they're going to change the world in unpredictable ways.  And for non-technical people that can be weird and scary because who knows what the consequences are going to be and who knows how to think about these things.  Money is interesting in that context because money is something that I would argue has been on the non-technical side of the culture for the last thousand years, for the last three hundred years for sure with paper money.  And then before that, which is when sort of normal people think about money they think about money in terms of who can you trust.  Who can you trust, right.  So which institution can you trust?  Can you trust the U.S. government and government officials?  Can you trust the U.S. Treasury Department?  Can you trust the U.S. Mint to catch counterfeiters?  It's very much, you know, can you trust the banks?  Can you put your money in the bank?  Are you going to be able to get the money back?  Do you trust the bank?  Do you trust the government standing behind the back?

And so our cultural, broad cultural view of money as it's based on trust in people and institutions, on the engineering side of things people in institutions are beside the point, the core of it is: can you trust the math, right.  Can you trust the math?  Can you trust the algorithms?  Can you trust the code?  And can you trust cryptography?  And so part of what's happening right now is engineers look at this and they're like, "Oh, it's obvious that Bitcoin is better because it's math that you can trust as opposed to people who you just never know" as we discovered in 2008.  Normal people look at that and they're like, "How can you trust math, like, does that mean you have to be a mathematician, right.  You have to have a PhD in mathematics in order to understand computer science to understand this stuff.  And so, I think what has to happen is actually the cultures have to come together.  And for that to happen basically, two things are required which is one is the engineering culture which we represent and which a lot of us here today represent.  We have to really work hard to be able to explain this stuff and articulate it and simplify it and make it easy to use and make it understandable, and we have to reach across and we have to bring people with us as opposed to just having this be something that's going to freak everybody out.  But people on the other side of the culture, people in the literary liberal arts, sort of non-technical culture, need to be open-minded and need to be willing to engage and understand and learn.  And that's a big part of what I think has to happen over the next three years.

Sheryl Sandberg:  Does that have anything to do why you've made the big return to Twitter?

Marc Andreessen:  Yes.

Sheryl Sandberg:  Being able to talk a lot about Bitcoin and kind of stirred the conversation?

Marc Andreessen:  Yeah.  I mean, collectively, I think we all of us, everybody in Silicon Valley, everybody in the Bitcoin world, everybody in the technology industry, I think we have an obligation to explain ourselves.  I think the days -- and again, this goes right back to C.P. Snow the days, where you could have something is fundamentally important as Bitcoin that just gets created, unleashed on the world, and everybody's expected to adapt to without having it get translated from weird and scary to, like, okay, you can actually wrap your head as a normal person around this, like, that's our responsibility, that's not going to happen by itself.  By the way that that had to happen with the internet.  It did happen with the internet.  It had to happen with PCs.  It did happen with PC.  So this does happen.  We're just right in the middle of it, right.  We're actually right at the beginning of it right now and so we still have quite a ways to go.

Balaji Srinivasan:  The other thing is also like, kind of, how this arose is not the way that, like, sort of it rose in the opposite of like an institutional top down kind of thing, right.  Bitcoin arose because somebody hit enter on the internet, right, like five years ago somebody enter in the internet.  They put something on a foreign post.  Now we're holding here today and you go to and all these governments are reacting to this person who basically hit enter on a foreign post, right.  And so that's completely different from the model of this top-down 20th Century model.  That's I think very common there.  One thing that we do have high hopes for are some of these new media institutions like FiveThirtyEight like  FiveThirtyEight news doing some stuff on Bitcoin and they are run by quantities of people who are kind of a merger of these sort of two cultures and so have high hopes for the kind of journalism that they'll do on Bitcoin.

Marc Andreessen:  And, of course, Forbes is still an outstanding.

Balaji Srinivasan:  Yes, that's right.

Marc Andreessen:  Must be noted.

Balaji Srinivasan:  Yes.

Sheryl Sandberg:  Thank you, Marc.

Balaji Srinivasan:  Right.

Sheryl Sandberg:  And, Marc, you -- I mean, you came forward at that very influential New York Times piece about why Bitcoin matters.  And I assumed everybody in this room has read that.  I don't need you to go over the points you made, but there was a response from Glenn Fleishman on media and he went into some of the negatives.  He said that you didn't cover including the fact that we don't know what will happen with transaction fees as Bitcoin continues to evolve.  I was curious which of Bitcoin's negatives you guys are most concerned about or central on your radar?

Marc Andreessen:  So conceptually, what is so striking -- and this is why we refer to Satoshi, I refer to Satoshi as a genius, he, she, it or they.  But I wouldn't rule out yet.

Balaji Srinivasan:  Interesting.

Marc Andreessen:  Yeah.

Balaji Srinivasan:  Yeah, a time traveler like Marc.

Marc Andreessen:  Software, software.

Balaji Srinivasan:  Yeah, yeah.

Marc Andreessen:  Robots and those Skynet.

Balaji Srinivasan:  Yeah.

Marc Andreessen:  But I refer to Satoshi as a genius.  And the reason I don't use that word, I don't use that word a lot, and I use that word very specifically which is -- and this is part of the adaptation process at least I went through in learning about this, which was you hear about Bitcoin and you're just like that's just crazy, made-up, fake mathematical currency like, you know, whatever, like the crazy kids are going at it again.  And then you get into it and you start to wrap your head around the peer-to-peer thing, and you start to wrap your head around the proof of work thing, and you start to get into it and you're like, "Wow," like that's a genuinely new thing.  I could build on a research that people been working on for a long time, but it's been brought together in genuinely new way.  And so that's sort of phase two.  And then phase three is you kind of come up with these lists of objections, right.  And how is it going to scale and how's it going to perform and what about this ten-minute thing and like isn't the Blockchain going to get too big?  And notwithstanding Newsweek's forensic expert disk space actually still does matter in 2014.

And so how are you going to deal with all these practical realities?  And then you dig into it and what you realize is at least every single issue I've been able to come up with Satoshi anticipated.  And this is really remarkable.  I mean, this is what is kind of so amazing.  And so from a conceptual standpoint, I think the answers are already present.  Now, we have a lot of implementation to do, right.  And so the ideas are there.  So, for example, transaction fees are great examples.  So the critique at Bitcoin is sort of what you said, which is like "Okay, today, you've got these miners who are like processing all these transactions because they have this self interest because they get these new Bitcoin coming out of the mining process."  What happens when that's no longer profitable or what happens over in the long run when the curve flattens out and there there's not a lot of new Bitcoin getting issued?  What will be the incentive to process all the transactions?  And then it'll be transaction fees and then basically the sort of argument goes and then you're right back to where you are today.

And that's where then you would go read the original paper.  And you're like, "Wait a minute, he thought of that."  And what actually happens is people have the ability starting now to be able to attach arbitrary transaction fees to transactions.  And even today, if you're a miner you can choose to mine transactions, process transactions that have transaction fees attached if you want to preference the ones we don't.  So there's already the beginning of a floating liquid market for transaction fees.  As you scale up over time, transaction fees will come to dominate.  But unlike today, where transaction fees are set by central institutions and by governments, in this market it will be a completely floating transaction, completely market-base transaction fee system.

And so the transaction fees over the long run will basically be the marginal cost of mining, plus a little bit.  And that little bit will determine how fast transactions get processed.  And so the thing that I know just like logically is there will be transaction fees and they will be low and the price will be driven down overtime because the system is wired to do that.  We all still have yet to build all of this, right.  So, most miners today are not prioritizing this.  There's a lot of work in the mining code that has to be done to be able to do this scale.  Most wallet services don't yet know what to do with transaction fees or don't deal with transaction fees.  And so there's a lot of implementation work to be done, but conceptually the thing is as bulletproof as anything I've ever seen.

Balaji Srinivasan:  And so building on that, in the early internet, you can find endless quotes, the one like embarrassing people who gave them who said, "Oh, the internet can't possibly scale to handle as much as people are talking about, how are possibly going to do online video chat when we can barely send a packet back and forth, you've got modems and so on so forth?"  They were right based on the snapshot of where they could see and like three years out, but not ten or fifteen years out with wireless broadband and stuff like that.  That's one.

Then number two is in terms of specific issues, you ask like, what errors we see for improvement?  So I previously mentioned like infrastructure, right.  So, but what does that mean?  So that means developer documentation, right, like that Bitcoin protocol should be the most highly annotated and documented thing out there.  I think grammars in standardization.  That might be too early to do now, but soon you're going to want actually like a standardization process sort of like the organic web standardization process of people who talk about the Bitcoin protocol, any new implementation of it would have to be book for book compatible with the old one.  We've kind of already seen that.  It's called web browsers.  We had to deal with that.

And I think BTCD, some of the new clients that are out there, BTCD is like run and go.  Some of these clients, it would be really great to start thinking about a standardization process.  That will happen after we've documented and what not.  So I think getting people into the actual core protocol and messing around the building and messing around the code is very important and investing more on that.  Now in terms of fees, one way I think that's important to think about is you've got the existing financial system, right.  And the existing financial system has ATM fees and it has overdraft fees and it has wire transfer fees and it has fees of every kind, both at the retail level as well as obviously at the investment level.  So you go into your IPO roadshow and you get investment banking that is itself kind of a fee, right.  And this is because there is kind of these super nodes that have certain permits and control over here.

And with Bitcoin, I think it's interesting to make a table of every fee in finance and ask which of them are actually going to be around in 20 years, right.  And how much of the profit will kind of go out of financial system like this, will kind of shrink like this, and then will be pushed out to businesses of the nodes and treat finance almost like a communications network, right.  And as an analogy to this, you think about the post office, right.  Okay, so you have a stamp to send a piece of paper mail, right.  I think there used to be more businesses that were actually run by mail, and then you take away the profit out of, you know, saying a stamp and now you can send an email for free and you can just send email free maybe.  Okay, now you can do like online courses.  And, in fact, you can do peer-to-peer online courses.  There's new business models that are not based on having this be the scarce resource, but you start have things happening at the end of those.  So with transaction fees, I think an equilibrium will be reached where it's enough to pay for miners in the future just like kind of supply and demand.  But I think that that's almost missing the point.  The point is turn finance intercommunication where we take the profit out of a lot of finance in the medium to long term.  And then now we have the profit at the end of nodes, people who are actually building businesses and doing things of value.

Sheryl Sandberg:  And part of the promise of Bitcoin isn't just the Bitcoin as a currency, Bitcoin as another way of transferring value.  It is actually this idea of crypto ownership and being able to use a Blockchain as a way of proving that you have something so it could be if something where the Blockchain evolves to replace Social Security numbers and you have a private key for your identity rather than this string of numbers that anyone else can copy.  To what extent are you guys looking at where the Blockchain is going to go in terms of the companies that you want to invest in and the way you want to see this evolve?

Balaji Srinivasan:  Yeah.  So, I mean, one way to think about it is we've got two billion mobile phones out there and one of the musings of mobile phones is even in India or Africa places that don't have the built-in phone lines with wireless you can kind of put a mobile phone in somebody's hands, right.  And so interesting medium to long-term use case, Hernando de Soto has written about how the poor in many countries don't have like defined property rights, they don't have things where they can borrow against and other things that they can say they have title to, right.  So at Bitcoin, you can sort of parachute in over a mobile phone eventually not right now but in the next several years, rule of law as a service, right.  So you can show that you have title for this or that object that you actually own it.  And it's already websites like, which will put notary public type stuff into the Blockchain.

And the question about whether you deploy something on the main Bitcoin Blockchain or as its own Blockchain and so on is going to, I think, to be depended on workloads and things of that nature.  It's sort of like you develop as a web app or a mobile app.  I think it's going to be workload dependent like Namecoin or DNS resolver may have a different workload then a payment system.  And so I think there's going to be decisions that are made as to, you know, technically which one is better.  But the Blockchain itself is a very interesting data structure where it will be abstracted out and used for lots of other things.  Already we're seeing a lot of pictures on, for example, proof of storage, or other kinds of proof rather than proof of work that are kind of taking ideas from a Blockchian and then remixing them.  Marc, yeah?

Marc Andreessen:  Yeah.  So, yeah, I mean, we're talking earlier about our investment fee.  So it's like I would say we think that like the Blockchain is the big deal and that currency is -- they are sort of BTC currency.  It's like one of a thousand applications.  It's the one that gets all the heat because it's the one that gets everybody else fired up, but the other nine-hundred-ninety-nine are going to be just as important or more important in the long run.  And so what we would like to do over time is basically a half, there will be a whole sequence of how these applications all get discovered and developed and then how there would be a whole sequencing companies that get built, digital title, digital ownership, digital media assets, digital stocks and bonds, Blockchain-based crowd funding insurance contracts.  There's a very, very, very long list of things where if you can have online trust in the way that the Blockchain provides.  You can reinvent, I think, field after field after field.  And that's a big thing.  That to us is probably even more important than anything that is sort of related with the currency today.

Sheryl Sandberg:  When do you think that happened?  What is the trigger for that where you switchover to this kind of expansion of the Blockchain?

Balaji Srinivasan:  So, I mean, it's is like sort of already happening in the sense that you can redeem Bitcoin for -- so, anytime you can redeem Bitcoin for some other kind of good.  It's almost like that good is being traded into the Blockchain right now, like, it is a very for sort of hack, right.  But I think that once people start putting, for example, software licenses or things that you can trade in there.  The exact first application is going to be pretty important that people use it for other than BTC itself.  Namecoin could be interesting in the sense of -- if you trade BTC for NTC.  You're basically trading a domain or the ability to buy domain in the Blockchain.  I think probably the next five years, certainly, we'll see it, probably much sooner than that maybe in two years.

Marc Andreessen:  I think if this technology had existed.  One of the thought experiments I was trying to do is, say, if this technology had existed at some point in the past, how would things be different today?  And if this technology had existed 40 years ago I think for sure DNS would be totally different.  And then quite possibly TCP/IP itself, assignment of IP addresses, for example, grounding protocols would be different if you knew you had this capability.  And then for sure all the application layer stuff.  So email would be different.  The web would be different.  All these social networking would be different.  All these things would be -- e-commerce would be completely different.  Digital media, business models would have been completely different.  And so I think you basically just -- the way we think about is just do the thought experiment of, like, "Okay, how would it be different?  Had you known that you had this technology back then, which you didn't?"  And run that in your head as kind of an x clean sheet of paper experiment.  And then basically -- and this is where I get aggressive, but then basically say, "Okay, then the world is going to reshape itself in that way."  In the fullness of time in the long run that is what the world will look like because it will be impossible to hold back the implications of the technology.

Sheryl Sandberg:  So if you run that thought experiment on one of those things, what does it look like?

Balaji Srinivasan:  So advertising is very good example, right.  So Chris Dixon and, you know, we've been talking about how advertising is sort of a way that -- because we didn't have micropayments within the internet, we sort of contorted ourselves into a way of the advertising to pay like one fraction like 1/1000th of a dollar when you visit a page based on your attention, right.  And then some fraction of people will pay.  But you can imagine ad-free websites that include Bitcoin as payments and you build basically Bitcoin wallet into the browser.  And just like every year you go and you refill your Google Drive with $20, you put your credit information.  You don't really think about it too much.  It refilled $20, $50 worth of Bitcoin each year and that's your browsing for the whole year, right.  And then if you have AdSense then Google has integration, so now you can basically pay with from Bitcoin and no longer see ads and you have a faster browsing experience, right.  So it's just like one example where we can replace potentially advertising micropayments with the real thing.

Marc Andreessen:  DNS, I mean, you have decentralized -- you've had on the internet this whole time, you've had decentralized DNS servers, decentralized companies that have become very big companies on the basis of running DNS as administrative domain names.  And if we had had the Blockchain 30 years ago, that would have never happened.  That would have never existed that way.  And by the way, the internet would be more resilient as a result because you wouldn't be able to take it.  You wouldn't.  One first thing that the government of Turkey decided to take on Twitter was take down his DNS record.  That would not be possible in the alternate world.  Now the Turkish government did discover that that wasn't sufficient and that they also had to take down the IP addresses.  But you could even imagine reinventing routing in a completely different way, a completely different peer-to-peer oriented way.  And again, it's not that all this stuff -- these are big existing systems, advertising markets begin to establish.  It's not like these things all get replaced in the next five years.  But give us 10 years, 20 years, 30 years, I do think the world will increasingly adapt to what's possible.  And we all have the opportunity to build the services in the companies that are going to make that happen.

Balaji Srinivasan:  I think a good rule of thumb is any system where the profit depends on being a supernode of some kind, that is a node that can do things that other nodes cannot is going to be changed, right.  So, for example, GoDaddy or anything is based on domain and it has a particular ability to read and write to DNS, that not everybody has, right.  And like a Wells Fargo Bank as some of those properties.  I think it's going to be a long-term process, but in general any vertical that has supernodes are going to be transitional intermediate forms and eventually you can have a system that doesn't have such dependency on supernodes.

Sheryl Sandberg:  Are you seeing anything yet of Bitcoin enabling new kind of deploying of these technologies?

Marc Andreessen:  We are seeing many new things.

Balaji Srinivasan:  Lots of things.  We can't --

Marc Andreessen:  Yeah.

Balaji Srinivasan:  -- talk to all of them.  But, yeah.  It's only advertising like micropayment-type stuff.  There's advertising replacement-type stuff.  There's a lot of things out there.

Marc Andreessen:  Yeah.  A lot of the stuff we're talking about today people are working on --

Balaji Srinivasan:  Yes.

Marc Andreessen:  -- they haven't announced.

Balaji Srinivasan:      Yeah.

Sheryl Sandberg:  Balaji, we talked before about Appcoin.

Balaji Srinivasan:  Yeah.

Sheryl Sandberg:  Can you talk a little bit about what Appcoin is and how it is?  Is this kind of the Blockchain protocol and part of Bitcoin?

Balaji Srinivasan:  Yeah.  Sure.  So this is like a post that, you know, Naval Ravikant of AngelList and I kind of worked on.  But so essentially the thought is that Bitcoin have potentially a way to rethink how we fund and monetize and even exit potentially, a company or an organization that there would be that a variety of protocols potentially can be rethought of as coins, right.  So DNS coin and Torcoin and PKI coin and so on and so forth.  Many of these protocols that have been backdoored in various ways by three-letter agencies could be reinvented in this distributed fashion.  And historically, the issue has been how do you get really competent software engineers to kind of do this?  Yes, open-source is being part of that and it's done extremely well.  But now you can have on top of that the bill need to say, have 30% of Bitcoins initially, almost like an equity stake reserved for the open-source developers and then you pay a small fee in that coin currency to use these servers, right.  So you distribute the code open-source and then anything that is in some degree it stand up at DNS coin server or a Torcoin server somebody have to send you some Torcoins in order to use that server, right.  And that would give value to it, right.

So that is a way to think about at least protocols, a variety of new protocols can now be rethought of and monetized and funded via quantification.  So, we could fund it by, for example, giving people $5 million for some percent of the coins, right.  So it's kind of one.  But it's more broader than that, right.  So it's not just protocols but, for example, namespaces, right.  So another way that you can think about, for example, monetizing a TwitterCoin like a distributed Twitter that'd be built on this.  So the server doesn't just have to be a DNS server or a Tor server could be a Twitter server or a TwitterCoin server.  And one way you could monetize it is or whatever, be a very valuable name.  And then you could have those names be rare and scarce and paid for in TwitterCoins.  And so Namecoin is kind of also like this, so you can extend it from protocols.  You can think about namespaces.

And really interesting questions, first, that those two right there are very, very broad ways to monetize things.  And I think we'll see dozens, if not hundreds of these kinds of things out there.  So Bitcoin is just the first in the list.  But I think what's really interesting and open question is, just generalize even further beyond that, right, if software is eating the world.  Many many services, many many things can be put into one of these two forms, basically, a web server that you're sending up and then paying in coins with that.  It's a substantially and extremely broad concept.  And it means, for example, a lot of things that are served on juristically right now, like valuations and things like that, can change because now you've got like a public ticker sort of for each of these things based on, kind of like to send me the Bitcoins market price to arrive that, supplying demand for these services, right.  So there's a lot of really interesting aspects of the Appcoin concept and that's something that we're very interested in.

Marc Andreessen:  And we have still funded one.

Balaji Srinivasan:  Yes.

Marc Andreessen:  I can't say which company.  We have funded one company --

Balaji Srinivasan:  Yeah.

Marc Andreessen:  -- that basically has this as their approach --

Balaji Srinivasan:  Yeah.

Marc Andreessen:  -- including their monetization approach.

Balaji Srinivasan:  Yeah.

Sheryl Sandberg:  Okay.  Interesting.

Marc Andreessen:  So we found one.  Or we would love to do more.

Balaji Srinivasan:  Yeah.

Marc Andreessen:  So we're in the final two minutes here.  So I was hoping we just end by -- if you guys wouldn't mind addressing -- Warren Buffett recently made this comment that Bitcoin is a mirage and encourage investors to run away from it because he said essentially that it's a great way to make new payments, but you can't make money on Bitcoin itself.  So I was curious if you guys --

Balaji Srinivasan:  Bitcoin has outperformed Berkshire Hathaway by a lot over the last year.

Marc Andreessen:  The historical track record of old white men who don't understand technology, crapping on new technology, is I think at a 100%.

Sheryl Sandberg:  Awesome.  Okay.  Thanks, guys.  Great chat.  And anything else you guys wanted to say to wrap this up?

Marc Andreessen:  How can we possibly top those two answers?

Sheryl Sandberg:  Now those were pretty great.

Marc Andreessen:  Thanks.

Sheryl Sandberg:  Okay.  Excellent.

Marc Andreessen:  Good.

Sheryl Sandberg:  Thank you, everyone.

Marc Andreessen:  Good.  Thanks, everybody.


Written by Melvin Draupnir on March 26, 2014.