Video - Bitcoin Security - Bubble Boy and the Sewer Rat

A whimsical security analysis of the open, public and decentralized Bitcoin system, compared to the closed, proprietary and centralized alternatives. Andreas managed to elaborate the significance of Bitcoin's security system and the decentralization protocol.


Andreas: Hey, everyone. Welcome. Are we having fun at DevCore? This is so, so amazing. One of the things I love about this space is that I just keep learning, like every single day I learn something new about Bitcoin and at least once a week my mind is blown. So I had that experience again today, listening to some of these presentations. I wanted to talk about security today. And if you listen to the trolls on reddit, I don't know anything about security, so I decided instead I'll talk about parenting because I don't have any kids. So, you know, if I'm going to talk about things I don't know I might as well start there, right. You know, parenting has changed a lot. When I grew up, things were very different.

Now in the last couple of decades, parenting is completely different. My sister just had a baby and I'm watching her as a parent. I'm like a proxy parents as an uncle. It's really strange. I'm watching these parents and, you know, when I was growing up, Purell didn't exist. It's a miracle we actually made it, right. Like, we survived because apparently these bacteria everywhere. And today's parenting involves gallon jugs of Purell, right. You watch these parents like their kid touches a bit of dirt and they give them a Purell shower right there just to make sure. Not the experience I had, right. I grew up in the '70s. We used to play in the garden, roll around in the mud, we'd make mud cakes, would our parents freak out? No. We'd eat the mud cake, would our parents freak out? No. Mostly because they weren't around. They were like get out of the house, come back when the Sun goes down. And so you have to wonder, how did we survive without Purell? And recently if you read some of the studies you hear about this really interesting phenomenon: the rates of asthma and allergy are through the roof. Turns out if you raise a child in a sterile environment, they don't develop an immune system. Oops. And so now there is this new round of parenting that is recognizing this fact that we're going back to our roots.

So now we realize that eating mud cakes in the garden is how you build a robust immune system, right. You don't get allergies, you don't get asthma. And, you know, you can take this to the extreme. You have, for example, in the third world, children don't have extreme allergic reactions to common medications that we'd have. Why? Because they have even more robust immune systems by being exposed to pathogens all of the time from the moment they're born -- before they're born. And then in the other extreme, you have this concept of raising a child in a bubble. Bubble Boy, right. You remember that story, Bubble Boy? It's a tragic story because it's true about a child without an immune system and there are these strange cases or medical tragedies where either children are born with compromised immunity or they lose their immunity through some kind of problem. And then they live in a bubble. And you have to be wondering, "What the hell is this guy talking about right now? I thought this was going to be a talk about security in Bitcoin and who we are? We're talking about bubble boys and eating mud cakes." There is a point to this. Hang on, hang on. So the reason I'm talking about this is because this has some really important implications in security. You see, if you create a system that is isolated from external influences then it's not that it doesn't have bugs. It's just that you don't know about the bugs that the system has. And if you create a system that is exposed to external attacks all of the time, it's not that it has a lot of bugs, it's just that you know about the bugs that it has because you keep finding them. And in the process, you fix them. And in the process, the system gets stronger.

So this all comes out of a discussion I want to have about an interesting phenomenon we have now, which is this concept of permissioned ledgers, an isolated blockchains. Because in my mind, an isolated blockchain is Bubble Boy, right. It's building a system completely isolated from the world with the hopes that, that's going to make it safer because banks are like a paranoid helicopter parent, the ones to shower their kid in Purell because it touched a booger. And guess what these lectures are going to get, they're going to get asthma and severe allergies. The worst case is that eventually the bubble bursts. At some point, you get exposed to the outside world and then you have a scenario where a system that's been isolated for so long has developed no immunity whatsoever. Its gets exposed to some horrific deadly thing like a pollen particle and dies a horrible death. Because it has such low immunity that it reacts horribly to something that a properly stimulated, properly raised organism can resist with ease. Now this isn't the first time we've had this discussion. In fact, ironically on the Internet this realization that security by isolation and security by obscurity and security by controlling perimeter and security by trying to tamp down security research fails and fails miserably.

When I was first on the Internet in the early '90s, I was talking to banks and telling them why they should get e-mail servers and connect to this e-mail thing. And they said many of the same things that I hear in Bitcoin today, which is well, "We don't know anyone who uses e-mail. None of the other banks use e-mail. So who am I going to send the e-mail to first place? Secondly, that out there uncontrolled thing might be dangerous. Thirdly, our bankers might say something in e-mail and how do we add a long disclosure form at the bottom and what happens if any of our people can communicate with anyone at any moment in time, that's a recipe for chaos, anarchy." Of course, they were right. They just didn't think of chaos and anarchy as a good thing. Many of us in this space probably do. So what did the banks do with their first attempt to join the Internet? What did large corporations do with their first attempt to join the Internet? Did they connect TCP/IP systems directly to the Internet and build robust applications that could communicate over TCP/IP? No. They built moats and walls and perimeters. They implemented perimeter security, they built firewalls, and demilitarized zones, DMZs. And they use all of these military analogies to wall themselves in. And then, what did they deploy behind these walls? Did they deploy the common open source protocols and capabilities and applications of the Internet? No, they deployed highly denatured weak equivalents, like, Outlook and FrontPage. And they built Internet websites that had stale and obsolete content that was only accessible during working hours through a VPN with no influence from the outside. And they said, "Look, we're doing Internet. We're so cutting-edge, we're hip."

And that's how they did Internet, they built these highly isolated environments. And for very long time, the prevailing idea was that by building these isolated environments they were more secure because they could control things through the firewall, because they could control access to data, creation of data access to systems. And now we know that wasn't illusion. Not only can companies not control these things, but in the process of building these isolated systems they built Bubble-Boy IT, they built IT systems that had no resilience, no immunity because Outlook had bugs and FrontPage had bugs. It's just that they weren't tested on the wild Internet very often because a lot of the time they lived behind walls. And when we discovered those bugs it was bad, right. Because eventually someone gets inside the bubble or the thing that's inside the bubble gets outside the bubble. See the problem with bubbles is that you can't trade through them. And if you're in business, your business is to trade. So if you're a business, you do commerce. And commerce can't happen in a bubble.

So the very concept of a bubble is antithetical to commerce, you build your firewall. What's your salesperson going to use on the road? A laptop, which they're going to take outside of the firewall for the very first time, plug it into the hotel Internet, contract 72 viruses, and then bring it back into the firewall, and give it to everyone else. Bubbles didn't work. On the Internet, it didn't work. What are we seeing now? We're seeing a whole generation of companies come to the realization that in order to be nimble and effective they can't be HPEMC, Cisco, Oracle Microsoft, Havens of secluded little kingdoms that don't talk to anything else. First of all, because that shit is expensive and it doesn't work. And secondly, because it is credibly vulnerable, it doesn't have immunity. And so now we see this generation of nimble young startups that are true Internet companies. Their products, their internal systems, their collaboration, all of it is out there naked on the Internet. It all happens on GitHub for all the world to see. They use Gmail and collaborate with external e-mail systems all over the world. Their internal systems are external. There is no such thing as internal in the world of the Internet. And they're building robust applications because on day one those applications live in the wild and they're more secure. They learn to live out there in the big scary Internet and those companies are thriving. And they have systems that are much more secure and much more robust. And that was even before the era of whistleblowers and anonymous who come along and prick these corporate bubbles and get inside and take all of the information and give it out. Now you're probably thinking, "Well if permissioned ledgers and closed Internets are Bubble Boy, then the wild Internet and Bitcoin are like a kid eating mud cakes," right? A system that has immunity, something exposed to pathogens. Well, almost. That might have been the analogy I wanted to go for, but you know me. I'll go a bit further.

Bitcoin isn't the kid that eat mud cakes. Bitcoin is a swarm of sewer rats, gnarly things, missing eyes and claws and tails like those pigeons you see in Trafalgar Square that are hopping around with this mutant arm stump. And what do they eat, what do they eat? They eat raw sewage, they eat your trash, they eat the most virulent things on the planet. There is nothing in this world that has more strength in its immunity system than a New York rat or pigeon or even, God forbid us, squirrel. Those things are horrible. And so a rat is not going to have allergies. It's not going to sneeze because of a bit of pollen. This thing is already carrying three variations of the plague and it shrugs it off, and that's exactly what Bitcoin is. Malleability attacks DDoS out there in the open. Port A333, come and get me. And is anybody trying? Hell yes, everyone is trying for six years. The best of the brightest, the meanest and the most malicious are throwing everything they can at this deformed swarm of sewer rats out there. These six thousands nodes that are listening and God knows how many other nodes that are exposed to the vagaries of the wild Internet and it survives.

So what do the banks do? They're going to build bubble-boy blockchains. They're going to build permission ledgers. Do you think permissioned ledgers suffer from transaction malleability? Hell yes, they do. Do you think Altcoins suffer from transaction malleability? Hell yes, they do. They just don't get those things fixed, right. And neither will the permissioned ledgers. And that's just one of the thousands and thousands and thousands of bugs and weaknesses and weird exceptions and edge cases that we're going to find while living out there in the wild. And we're going to build this incredibly robust system, which is already taking shape today. I mean, beyond the idea that you could have a decentralized consensus system. The idea that that decentralized consensus system could actually survive for six years is kind of ludicrous. And the only reason the banks have now gone to the point of thinking about permissioned ledgers is because they finally reach the stage of bargaining. The third stage, in the five stages of grief, for the industry they're about to lose. They start with denial. And the basis of denial is, well, this thing isn't going to work. It's going to die any day soon, and it doesn't. And then they say, "Well, it's just silly money and it doesn't have any value" until it does. And nobody else is going to play with it except that they are. And serious investors won't possibly put money in this except that they did. And it still refuses to die. So we go from denial to bargaining.

Somewhere in between there might be some anger, there is going to be some depression, and eventually they're going to reach acceptance. But it's going to take a long time. Because if you look at the Internet, we're now on maybe 25 years into the Internet in terms of really beginning to broaden its use. 25 years in and there are plenty of companies out there that think that as long as they put their Oracle, EMCHP, Cisco, Microsoft shit behind a perimeter firewall, all is going to be well. They're still building bubble boys and intranets on the Internet. They haven't learned that lesson after 25 years. It's going to take longer in finance. Not only is decentralization, open protocols, open source, collaborative development and living in the wild. A feature of Bitcoin, that's the whole point. And if you take a permissioned ledger and you say, "Well, that's all nice. We like the database part of it. Can we have it without the open, decentralized, peer-to-peer or open source, non-controlled, distributed nature of it? Well, you just threw out the baby with the bathwater. You're never going to build a bubble strong enough to keep financial information. Ironically, this is all happening at the same time that as banks have finally gone onto the Internet, they're leaking. They're leaking so much from every orifice, they're leaking. Anonymous, WikiLeaks, Insiders, all of that stuff. They don't have confidential transactions. They don't have encrypted this. They don't have privacy. They don't have zero knowledge. They have completely open ledgers. And what do they overlay on top of them? KYC and AML. So they attach identities to everything they're doing so that when that database gets leaked it will have a completely rich history, not only of every transaction but of every participant in the system. That's what they're building.

They're building Panopticons. They're building a Panopticon of financial information and it's leaking. Because the truth of Panopticons is when you build a Panopticon it stares back. And when it's the Internet that's staring back, that's four billion eyeballs. I'm not so worried about my financial information from my bank leaking because maybe a couple hundred people are going to stare back. But when Angela Merkel's phone numbers and phone calls leak, woo, everybody's staring. Three days ago, the internal presentations and PowerPoint of the Department of Defense about their drone assassination program leaked, four billion eyes staring back. You built Panopticon, it's staring back. And so the real question we should be asking about permissioned ledgers is, "Do you really want to put KYCAML on Bubble Boy?" Because you go and add all of that information when that database leaks four, five, six, ten years into the future you're going to give anonymous WikiLeaks historians, a complete record of every transaction you ever did, the secret slush budget of Lockheed Martin, the black budget of your government, the bribes that you paid to depose a democratically elected government or to install an oil well and a pristine rainforest. All of that shit is going to be on WikiLeaks and all over the Internet. And you're going to provide the rich KYC metadata that you painstakingly attach to every transaction. Meanwhile, we're going to build Bitcoin with encrypted anonymous private transactions and you'd better rethink this Panopticon, you'd better rethink this Bubble Boy because building resilient systems is about exposing them, exposing them to continuous attack. That's how you build resilient systems. So I'm not scared of permissioned ledgers. Denatured, defanged centralized weak systems behind bubbles, those are not going to scale, they're not going to survive, they're not going to be secure, they're not going to be provide, they're not going to be providing privacy, and they're going to backfire badly.

But the funny thing is that lesson is going to take a long time to learn. I can see it now. "Sir, we had all of the drone assassination things behind a firewall but someone burst through the bubble." "All right. Call the general. Get me two bubbles. We're going to double up. Bubbles within bubbles." "Sir, they burst through our double bubble, titanium bubbles." "If we pay Lockheed Martin a hundred million dollars maybe they can build us a double titanium bubble that we can hide all of our data behind." "Sir, it lasted 30 seconds before anonymous ripped it to shreds and put all our data on the Internet." "Hmm. I wonder if we can build more bubbles." They think that having your data on the Internet without controlling it centrally is weakness. It isn't weakness. That sewer rat out there isn't weak. It's the strongest thing we can build because it's constantly under attack. And wrapping it in a bubble, it doesn't make it stronger. It gradually denatures and weakens it until what's left is a pale immunosuppressed little live rat with red eyes that dies the first time it's exposed to the flu. And so, that's what security is. Security is a process. It's a process of openness and exposure. It's a process of continuously adapting to new attacks and in that process dynamically becoming more and more robust, less and less fragile. We're introducing Bitcoin in a world full of fragile systems. Central banking, centralized banking, monetary systems that can't manage to achieve liftoff in the economy. In that environment, we're introducing a robust global decentralized system. And it's robust today. It's not perfect, it's got bugs. But we don't hide those bugs. We announce them. We glorify in them. We discuss them. We invite people to attack it and we take that information and we make it stronger every single day. And that is why we win because while they're building Bubble Boy, we're building a swarm of sewer rats. Thank you. So I'm happy to take questions from the audience. We have quite a bit of time. So please go ahead. Andrei (ph).

Andrei: (Inaudible 00:24:25). What you're trying to communicate is that rival blockchains are insecure by design. I mean, blockchains that are built within the banks. Okay. I agree with that. But we can take another software that is being used. I mean, as example, like open source project, okay, like HTTP server, or let's take an Nginx or Apache. It's been used by big corporations like Google, Oracle whoever, including banks, that have a lot of private information. And so what prevents the banks from taking open source grown copy of the Bitcoin code and launching it inside?

Andreas: Well, I'll tell you what, what stops them. I think here is the problem. What happens if you take Apache and you install it in a bank and you put it behind an intranet and you use it internally? I'll tell you what happens. You fall behind on the patches, you stop doing vulnerability tests, you stop exposing it to external vulnerability tests that you didn't order that just came your way.

Andrei: Okay. So that --

Andreas: And as you do that it gets denatured. It gets weaker and weaker and weaker and weaker until eventually you're running Apache. But it's three versions behind it, vulnerable to anything. And someone comes in, breaks through the bubble, breaks through the perimeter, and takes that Apache for a ride. And that's because you weren't under pressure to live in the wild. And when the pressure goes away, so do the standards.

Andrei: I would be happy to see Bitcoin as the one world currency and you probably know that I've also been working towards this direction during last five years or so. But meanwhile, we have big banks and corporations existing within the countries. And Google is a good example of like using a lot of open-source software and using it properly, right. Do you agree with that?

Andreas: Yes. And most of their stuff --

Andrei: So why all these --

Andreas: -- runs pretty much out there. It has to.

Andrei: Why don't we have not yet shifted to like completely decentralized anarchists like picture of world with only one currency? Will we have the banks? And Bitcoin solve some problem for them. I mean, private blockchain solves problem of synchronizing transactions between the branches, like, not losing transactions and so on. So they have a choice either to, like, not solve this problem or try to apply this solution?

Andreas: They have a lot more choices with that. I mean, just today Greg was talking about liquid which is a side chain for doing exactly that between exchanges. Now where are exchanges today? Today they run a MySQL database that stores entries for the account value of every customer. We saw what happened with Willy Bot and Gox with that particular issue, right. This is an incremental improvement. Now, how does that differ from a permissioned ledger? Well, the main difference is that if you think Citibank is going to run their permissioned ledger on Internet-connected machines and open to everyone to scrutinize, you're sorely mistaken. What they're going to do is they're going to hide it behind a tall wall and they're going to run it among their five, six, seven, eight bankly friends. And what that's going to do is it's going to mean that that software is going to be weak and it's going to get weaker because all of the lessons we're learning in the wild won't get applied there until the whistleblower runs a little Trojan and malleates the transactions of their running exchange. And then they're going to have a bit of a problem.

Andrei: Okay. I got you. Yeah, what I was trying to say is that as long as big institutions still exist, they will hold some amount of private information inevitably about their customers, right? And since blockchain solves some problem for them, they will be using it. And they really have a choice to either use open source developments like launch a copy of a Tyrian within their network or try to build something on. So we'll have wars of blockchains again. Like commercial blockchains built by Microsoft and open source blockchains built by open source community. And both will be used by large organizations. It's not like --

Andreas: Yeah, absolutely. I mean, we are going to live in a -- thank you. We are going to live in a world with a lot of diversity. We are going to have completely closed systems that are permissioned ledgers that have so little decentralization functionality that effectively all they are is three phase commit on top of a database with audit logs. And instead of having audit logs in a log file, they have audit logs based on Merkle trees and hashes. That's not innovation. That's 20-year-old technology applied in a slight twist to what they're doing now. And on the other end of the scale you're going to have completely open systems, open source systems, you're going to have sophisticated cryptography. And we're mostly going to be living on that end. Now if that's the environment and that's the competitive landscape, that's great. I mean, because that's an environment in which not only can we win with Bitcoin and with other technology. Or rather it's not a matter of winning, it's a matter of building robust solutions that have use and value for people all around the world that changed the world. That's something we can do. You know, I'm not worried about competing against the Microsoft blockchain.

Unidentified Male: You know, you described Bitcoin as our army of sewer rats. But I'm going to disagree, I think our Bitcoin is a single sewer rat. And because of that it's vulnerable. The sewer rat is named Bitcoin Core. If we really want to be army of sewer rats, we have to have more implementations so that if one rat dies, there are army remains.

Andreas: You know, I don't think you will find a single core developer, who will ratify the idea that the best approach is to have only one implementation. I think the real difference is that implementing more than one implementation in creating software diversity on a consensus sensitive system is something that's never been done before and it's bloody difficult because you have one bug.

Unidentified Male: I totally agree.

Andreas: And you get the May 2013, 26 block fork because of Berkeley DB which wasn't even part of the consensus rules. I think if you look at the development roadmap of Bitcoin Core, you'll see that there is an enormous effort underway which lift consensus and lifts up 256k to modularize and isolate the elements that are consensus important. And to make those available for libraries for other implementations, and there are other implementations. Year six is a toddler, right. And so already there are three or four competing implementations that are fairly good and are able to keep up in some ways. Is it still very much a monoculture? Yes, we still have some biodiversity issues. But I don't think anybody wants that. It's just they recognize that it's very difficult to move away from that in a system that is consensus critical. Okay. Question. Michael.

Michael: How much of it is the question of competence? I mean, looking at operating systems, iOS is a lot more secure than Android. And did that kind of breaks down your analogy?

Andreas: I don't think it's a matter of competence. I think, well, it depends on how you define competence. If you think of competence simply as an internal and intrinsic attribute of a single person then perhaps the competence isn't to me an attribute of a single person. It's an emergent aspect of a team or collaborative behavior, right. Competence is not you writing code alone. Very few people can exhibit competence across scale and time as coders. Competence and quality of code is something that emerges from the collaboration of many people because the area that I have competence in is different from the area that you have competence in. And if we're sharing then there will be someone out there who will notice the one thing that I missed. And so I don't think that's really the case. I think what is the issue with Android. I think the fundamental difference between Android and iPhone is not about code quality or security of the underlying code, it's about the difference that iPhone runs on 20, 25 different platforms if you take all of the versions of iPhones that exist out there. And Android runs on 500 different platforms by different manufacturers, all of which creates subtle variations. It's a matter of uncontrolled diversity in a system. And there is Android that is extremely good and there is Android that is extremely bad, whereas with iPhone it's a much narrower band of higher quality. That's a specific choice to align hardware software quality control services under a single umbrella. And that works in some cases, but it also slows down innovation. And how do I know that? I know that because I had a Bitcoin wallet a year and a half on my Android before it was available on iOS. And that's a perfect example of how it slows down innovation, wall gardens, mini bubbles. They reduce your ability to trade outside the bubble and so you pay a heavy price for that. And over a longer scales of time, that price may be insurmountable.

Unidentified Male: I love the sewer rat analogy. That's awesome. And the helicopter parent, that's equally awesome. Still trying to wrap my head around how the sewer rats see the helicopter parents and how they relate to them and whether they ignore them completely and --

Andreas: They live under the crib.

Unidentified Male: -- go grab the business. It seems like the common element of these semi cooperative entities and rats don't really collaborate and helicopter parents.

Andreas: I mean, listen, I'm not going to make a --

Unidentified Male: Somewhat elaborate, but --

Andreas: Okay. Let's not attempt to do a formal proof --

Unidentified Male: Yeah, yeah.

Andreas: -- on the internal consistency of my analogies. I can tell you right now.

Unidentified Male: I'm just saying I love it I'm going deep on it so.

Andreas: The point is not about --

Unidentified Male: I could talk about this whole day.

Andreas: -- this specific biology over the rat. The point is about the difference between robustness in an environment with stimulus versus weakness in an environment that lacks stimulus or has isolation.

Unidentified Male: Uh-huh.

Andreas: And so use whatever analogies you want. I thought that starting a title of a presentation with Bubble Boy and the Bitcoin, sewer rat, as I announced on Twitter last week, would at least brings the people here thinking, "What the hell?"

Unidentified Male: Absolutely. Thank you.

Andreas: But in addition, I think that control is the medium that -- or the perception of control is the medium perception of control.

Unidentified Male: And when you said rats, I immediately thought of pizza rat in the New York subway system and --

Andreas: Yeah.

Unidentified Male: -- the helicopter parents like Pizza Swizz or maybe pizza is the control.

Andreas: Well, here's the thing, the control is going to be a big issue with these permissioned ledgers. The illusion of control or the use of hierarchy authority and control in order to effectively change the future, that's an illusion that all of us can fall into, right.

Unidentified Male: Right.

Andreas: The assumption that we control our destiny and that if only we control a few more variables we'll have control. That's what drives people crazy, like, if you want to be neurotic. Now if we wanted to make the analogy that many large corporations are institutionally neurotic, I'm all with you because effectively that's what that element of control is. That's being terrified to open yourself up to the outside world because you are a hierarchical institution that has authority and control in its very DNA. And that being a fundamental and perhaps extinction-level weakness of large hierarchical organizations. I'm with you there because that is the end result of this. It is an issue of control.

Unidentified Male: The ability to infectiously (Inaudible0:37:31).

Andreas: Yeah, we're not infected. All right. Let's take one more question here. I think maybe we have a bit more time. Go for it.

Unidentified Male: Thanks a lot. Also could agree with everybody with cool analogies, but I just would love to hear your thoughts about whether Bitcoin has actually been attacked in all the ways or in the most effective ways. Because one thing that makes sense to me as, you know, if I for example had an attack that was very effective, right, I wouldn't use it now when I can't profit from it. I would wait until I could profit from it. Namely, when a short market appeared where if I successfully executed the attack I would make $10 million, $20 million, you know, any millions of dollars. So I'm just not -- I get the analogy and what you're saying makes sense compared to distributed ledgers. I'm just wondering on your perspective of what happens when much more economic incentive via short markets appears for somebody who might have an attack to actually use it.

Andreas: That's a really good point. And I think we should recognize, and let's be realistic here. Bitcoin has not been attacked in every way possible and as much as it possibly can yet and it certainly wasn't in the early days. Bitcoin had one unique advantage which was this two-plus year honeymoon period when nobody thought it was important or relevant or even would work. If at that time people had attack it, it was much weaker, right. There were some horrific bugs in the early days, right. And there are plenty of core developers here who could talk about some of the hilarious things like, for example, being able to create Coinbase with billions and billions of Bitcoins in them. Oops. Some of the validation rules slip through blocks that had infinite coins in them and many other bugs. We got a honeymoon period then to fix the most egregious bugs. And we still have a honeymoon period now because here's the hilarious thing, most of these banks, most of these large organizations in finance, most of the central banks, they look at Bitcoin the way Walmart looks at a lemonade stand. And they are still laughing, which is great. I hope they keep doing that for two more years, three more years. Give us a bit more of a honeymoon period so we can get even more robust. Because we really don't need concerted attacks right now, although from another perspective I would rather have some of the attacks materialized now before we have mass adoption and a lot of users being disrupted. But this is a continuous process and a race. The real issue here is the time scale, right. And the interesting implication what we're saying here has is that a lot of Altcoins don't get that grace period anymore, which is why it's a lot harder to build robust Altcoins because, one, you don't get a grace period on mining. If anybody thinks it's going to be valuable they're there so it's not just like nobody noticed. And you don't get a grace period on security anymore. So if you've implemented things sloppily, someone's going to find it. In fact, just the other day I was reading this fantastic article about 42-coin. Are you familiar with 42-coin? It's an Altcoin that was designed to only ever have 42 coins. It currently has 48. It would have taken two lines of code to constrain the mining algorithm so that after the initial process of mining the first 42 coins as promised it stopped. And in fact, several people notice that this was missing from the code and they wrote to the developer who had since abandoned the project. And so nobody patched it and nobody upgraded these systems because they were really running in an isolated environment and not really participating in a real economy. So nobody fixed them and then coin 43 was mined. And at that point, you have an existential crisis for this Altcoin because it's no longer 42 coin. This is going to keep happening and it happens because there's not enough people interested in fixing the bugs. You know, this is the flipside of this idea. It's really hilarious to me. When you talk to companies and you say "Hey, how about you open source your code?" And they say, "Oh my God. If we do that people are going to see it and they might use it without paying us." And the hardest thing to explain to a company that's doing software is you wish people would see it and use it. Most likely if you open source your code like the other 700,000 projects on GitHub, no one will give a shit and no one will use it and you will not create a community. If you actually manage to get people to see it, use it and create a developer community around it. "Congratulations, you're in the 1% of projects that have achieved that." It is a rare and difficult achievement. And, in fact, Bitcoin has succeeded more in that than any of the Altcoins, or to go back to my previous analogy, any of the permissioned ledgers would ever hope to have when they closed themselves down from external scrutiny. Do you want to pass it to the person next to you please? Thank you.

Unidentified Male: Okay. So I really like sewer rats too and I see them in New York City screwing around the subway and maybe they could survive a nuclear apocalypse or something more than the bankers up above. But they're also living in muck and dirt in these little small passageways while the bankers up above have huge buildings and they have also have a lot of power and they could go live in Bermuda. Their bubbles give them a lot of ability. I'd like Bitcoin to have a lot of power, too. How did the sewer rats get power?

Andreas: Well, here's the funny thing, 650 million years ago, there was a big lizard species or a series of big lizard species on this planet. And they were big and they were proud and they tromped around when they stomped around and they usually stopped them on little furry mammals that were scurrying among the tree trunks below them. They didn't pay much attention to them. But guess what, they died. And the little furry things became us and we won. So don't underestimate the little furry mammal among the trees because someday meteors happen. And here's the thing, when the dinosaurs see the meteor they go through the same process of the bank seeing Bitcoin. They look up and they go, "Well, that's not happening. That can't be real." And then they start screaming at it. So to me, the banks at the moment dealing with Bitcoin, the ones that have begun to realize what is happening are now braying at the meteors trying to make them stop falling on their head and you can't really do that. Don't underestimate the tiny scrappy little incumbent, the little competitor scurrying around the tree trunks, because eventually they become the dominant species. And remember where and how the Internet developed in the beginning because I went into phone companies and did presentations explaining to them why they needed to address and understand and adapt to the threat of decentralized communications. We didn't call it that then, but TCP/IP. And you know what they did, they laughed. They laughed at the Internet, these massive companies like AT&T. And in my case, I went to the Greek national phone company, OTE, doesn't really exist anymore, fell apart. But they laughed at the idea of the Internet because the idea that through this messy process of decentralized routing where you draw packets all the time, that's not a bug, that's a feature dropping packets. It's messy. It's nasty that this could actually compete with these carefully constructed hierarchical systems of these global spanning copper and increasingly digital networks. It was completely laughable. They went off and designed ISDN and said "Huh, better than the Internet. You can do videoconferencing." The Internet can't scale to do video conferencing or voice or any of those things. Fast forward 20 years now, they're running their entire voice network on top of the Internet. Things change much faster than we anticipate and the power and scalability of decentralized systems and the robustness of systems that initially appear to be messy and sloppy like the Internet was -- can often surprise. But what doesn't or shouldn't surprise you is the hubris of those who think that the little scrappy competitor won't amount to much. All right. I'll take one more question and then we'll wrap it up. Thank you. There you go.

Unidentified Male: Alan Turing in The Enigma demonstrated that no form of cryptography, any form of cryptography can eventually be broken all through history when you had Navajo and various types of cryptography. There was always something that nobody imagined that would crack it. Do you believe that to be true or not?

Andreas: Yeah, absolutely. All forms of cryptography can be broken. All forms of cryptography --

Unidentified Male: Including that book.

Andreas: -- are eventually broken. That is a truism.

Unidentified Male: Including that behind Bitcoin.

Andreas: Including that currently behind Bitcoin, yes. The question again is time scale. You see the real secret of The Enigma was the secret of the broken Enigma. The reason Bletchley Park was successful in essentially winning World War II at least for the North Sea and the British Forces was because they managed to hide the secret of breaking Enigma. Because what would have happened if the secret that they broke Enigma leaked.

Unidentified Male: Yeah.

Andreas: Enigma would have been improved and changed and the damage that they had managed to cause which at that point was complete and systemic capture of all of the cryptographic communications of the Germans would have been contained. And so they would only be able to capture the Enigma machines that hadn't yet upgraded. Isolation was the downfall of that system because by definition it had to be isolated. So the lesson we need to learn is we expect cryptography to be broken, we expect every system and subsystem within Bitcoin eventually to be weakened. And what we need to do is, one, make sure that any such weaknesses are not systemic and complete. And then identify the weaknesses early enough to start addressing them so that they don't become systemic. And the best way you do that is by existing in an open collaborative environment where you learn about those weaknesses. If ECDSA gets hacked today or becomes weak today, what does that mean? Does that mean that every person in the world can suddenly crack ECDSA at any scale? No. It will mean that for a certain class of very well-funded attackers, certain types of ECDSA with an enormous effort can be cracked. At which point our friend Greg back there we'll be building a side chain that doesn't use Secp256 K1, Secp256 K1? In fact, the example of the snort signatures implementation on elements Alpha already shows you the possibility of having a Bitcoin subsystem that allows for a variety of signature technologies to be used within the Bitcoin ecosystem. There is no reason why we can all need to use ECDSA. We can add a patch to the system that recognizes, let's say, Apple's curve that they used, I don't remember what it's called, it's a long number, or that uses a completely different cryptographic system. I probably select something, created verified audited by Bruce Schneier. But the bottom line is that you could create, in fact, an ecosystem where you don't rely on any single curve and therefore the system is robust because every customer can pick which curve they want to use or which signing system they want to use so that even if one of them was compromised that only compromises a subset that's possible to do today. The real question we need to ask is, like, two weeks ago, SHA-1 was shown to be weak. Eventually SHA-256 is going to be weak. And at that point, we had better have reached the point in the curve where fees matter more than rewards. Otherwise, the consensus mechanism won't let us upgrade. But they're always weaknesses. No cryptographic system lasts forever, which is why you don't want to bake it into a permissioned ledger behind the wall that nobody ever inspects, maintains or updates because then it's going to become weak. And in fact, those systems are going to become monocultures. They will lack security biodiversity to use the term strangely, but they will lack the diversity required. Bitcoin is not very diverse today, but it is getting more diverse and will continue to get more diverse and more robust. All right. Thank you all. Appreciate your time and thanks so much for coming.


Written by Andreas M. Antonopoulos on November 10, 2015.