path: root/transcripts/bitcoin-devcore-2015/jeremy-allaire-circle.mdwn

Circle

Jeremy Allaire, Sean Neville

We are excited to be here and sponsoring this event. We have backgrounds in working on developer tools that goes back to the early days of something.

How do we mature the development of Bitcoin Core itself? One of the things that is useful is suss out the key components of it. In a standard you have a spec, it could be a whitepaper, and then you have a reference implementation, and then a test suite that enforces interoperability. The test suite is what enforces the standard. It's not the text in the whitepaper that says what the standard is... If there is anyone that can satisfy how many hundreds of thousands of tests, if you can write software that can satisfy these tests has written an interoperable implementation. The existing C++ code, there's the Satoshi whitepaper, but really the code the source code itself is the spec. The RI is important. They are explicitly tied. I am not talking about certification testing or branding or licensing.. if someone was crazy enough to reimplement this ledger and the set of protocols, that it's actually interoperable with what has been built in Bitcoin Core, there is no test suite for that today... How do we mature the process? If I were a developer, what is the first pull request that I could submit ? I would start with the test area. And then I would do it with the thought that ultimately what needs to happen is there should be some sort of test suite that ensures that when new variations that are introduced that there is agreement that what the interop looks like. And then that allows other people to get plugged into the process, there is no big committee that meets, there's no giant software companies that have a voice, it's more a meritocracy which I think is okay up until today, but sometimes when projects mature it becomes difficult for people who do not necessarily fit within that culture of communicating in this particular dev email list or on IRC, people who have nothing to add don't really fit into developing software. So it's more like how do we pull people into this, how do we get software companies into this system? What's theright format without getting overly complicated? How do we let people plug into this system? How do we make it easier for people to write code and innovate on what we're collectively building?

"Trust".. what does that involve? Is there a generational difference in how people think about this technology? What do traditional institutional stakeholders need to trust to integrate established industries with this technology? How do we think about accountability which is part of trust, with the kind of stakeholders that are involved in the infrastructure layers of Bitcoin and the blockchain? Our target market inherently trusts internet technology. That it can do anything. As you get into older generations, they don't have that trust. They distrust those facets. It really plays out. The younger audiences assume instant connectivity. They think of it as a public good. The folks that have been running intermediaries have the opposite reaction. We interact with bankers, auditors, accountants, lawyers, regulators, who else that is big and scary? Those are the big scary guys. Just the idea of using math and algorithms versus using people is really hard for these institutions. If you step back, there's these trust industries that were built on control over rights to making entries into ledgers, intermediating access and claiming provenance and authenticity on those records. That's banks, law firms, accounting firms, auditing firms, insurance companies, mortgage companies, etc. In the future, we think that a lot of the things that people do with these manual processes about ledger and records keeping could be automated in interesting ways. This is scary to those people. It's a key issue. It's enlightening to me that when we were in DC, we got not grilled but we got lots of questions from people in regulatory environments it kind of realizes the key issue is that using cryptography is far more strange to them than trusting some individual to do things. It's not even really a technical issue, it's a big disconnect, you can kind of see as we're talking about various topics you kind of see people nodding their heads, those are the people who would be okay with using computers, whereas other people think it leads to the Terminator showing up and he will be adjucating our payments or something. We're regulated now, we're regulated as a kind of financial institution and by law we're required to keep full reserves of our customer's funds, whether dollars or bitcoin. In the past, if a company like us had to be audited, the auditor would ask for bank statements, or could we get access to a report that your Oracle financials produced, and we will trust that nobody manipulated Oracle financials, or we trust that nobody is doing different journal entries in the firm... When we tell them that you can monitor addresses, they can't understna dthat. (Well duh, that's because it's a false statement). They are far more comfortbale with an excel spreadsheet (then you should integrate vbscript/rpc commands, sigh). The person who has the job of stewarding numbers are terrified of this. There is trust of people in the system, the trust issue is complicated, we have taken away the need to trust people... lots of people don't even read the release notes when they upgrade hteir nodes, they don't read pull requests, so they trust that the process of open source development will result in software they can trust, when someone signs a release they are trusting the whole chain... they are trusting people to be incented by the incentives implicit in the protocol.. Obviously the protocol is setup so that people are incented to do the right thing, and we're trusting that people are incented by financial incentives at all. There are people who don't care about financial gain who would be willing to destroy the network. We are trusting miners to a certain degree (??). Not having to make decisions about transactional echange is huge.

"The root problem with conventional currency is all the trust that’s required to make it work. The central bank must be trusted not to debase the currency, but the history of fiat currencies is full of breaches of that trust. Banks must be trusted to hold our money and transfer it electronically, but they lend it out in waves of credit bubbles with barely a fraction in reserve. We have to trust them with our privacy, trust them not to let identity thieves drain our accounts. Their massive overhead costs make micropayments impossible."

"Identity is a key concept"... (blah, no thank you). Reputation, federated identity, regulatory approaches to know your customer, anti-money laundering, fraud, risk and things like that in transactions and it layers into protecting consumer privacy and avoiding criminal anonymity for certain kinds of things. It's an important issue. I have long been passionate about identity for 12 years including federated identity. There is a fundamental relationship between government identity and financial identity. They are tightly coupled. If you talk with governments and large financial institutions, these are the issues they are grappling with. They want identity. They want identity for their citizens. Sean and I have talked about these themes. I think the blockchain has blown open the reputation and identity stuff.... the concept of Sybil attacks, and public key cryptography and how people ((don't)) need identity is a really powerful concept. It's pretty important. There are two sides to when we talk about the benefits of Bitcoin as a transaction medium. One is that, hey, we have this existing internet which has been built on.... ((Nevermind, I really can't tolerate this. They don't seem to understand Sybil attacks.))

Complexity is a key problem in terms of Bitcoin adoption. This idea of a spectrum of control... having absolute control of private keys, and simplicity... The simpler you make it, the more control you have to give up is not true. This is intellectually lazy to think of this as a spectrum. I think both of these things could be accomplished at the same time. It's not about stripping out control, it's about the essential nature about what people should be able to use the blockchain to accomplish and then iterating on that. Complexity and mainstream adoption means something different to Bitcoin people than just mainstream consumers... we talk about reducing complexity, we're talking about workflows for multisig or securing your keys in hardware. Mainstream consumers don't care about any of that stuff. It has to be far simpler. It has to be like Steve Jobs. If it has a stylus, you have lost. There has to be another level of simplicity who don't want to join the Bitcoin movement but still stand to benefit from Bitcoin. Maybe they don't need to be aware that they are using Bitcoin, perhaps they get the near zero fees and near instant settlement. There needs to be a different kind of thought to address complexity for actual mainstream consumers. Otherwise we can make Bitcoin simpler. Really what we're doing there is finding the early adopters instead of actually moving to the next group which is the early majority and getting them on board.

Is 2015 or 2016 the years when we can go mainstream? I'm constantly asked by family and friends and people etc is what is the killer app. The killer app means introducing new layers of... it emerges when the technology disappear, when web search made HTTP become a non-issue. When you can sign up and use a web mail account instead of configuring SMTP. It's about making the technology disappear. What's the right token ring, to make this simple? If you are thinking about token rings, then you've already lost. The right abstractions need to be put in place. We're right on the cusp of the benefits, like near instant settlement, global interop, better security, better privacy, these are huge advantages over existing payment systems. To the average person it's still too much of a commitment. That really is a turn off to a lot of users. One of the other things on the complexity side, if it was just a problem of creating software, things would be easier for us, unlike in other businesses which have involved meaningful infrastructure and tools for other users, this is the first business where I need permission to write software or launch a product, I need permission from the government to do that. So there's another kind of complexity with Bitcoin itself which is it really you know to to deliver truly great experiences and to to make it work for users it requires sort of interop with the existing legacy payment systems and the existing ways that people move value to make it useful to peopel today and to make it useful today we can innovative on Bitcoin in a permissionless fashion which is amazing but we can't innovative on top of ACH or the card networks without a great deal of oversight and that makes it complex to get mass acceptance because we have to invest more and you could argue whether this is good or bad and does it hurt our ability to grow and that's worthwhile to debate. I think that is the basis for the constraints of blockchain adoption. Certianly lots of startups have felt the impact of that, like unexpected fees or things like that, which I think you know makes it very very different to other technologies that were made simpler.

Closing thoughts. Yeah. Just sort of. I think, high level, I think, Bitcoin is going to have profound impact in over 10 or 20 years. I think the impact is going to be far greater than we anticipate. There is this sort of truism. New technologies on the web in the 90s take lots longer than advocates think, but the long-term impact is often greater than what the strongest advocates think. We had giant ideas about web being available everywhere. Expansion will take longer than we think, but I think the long term impact will be more than we think. Fiduicary trust apps.. these kinds of apps touch every institution on the planet, everyone needs these intersections of accounting and payments, they weave their way through every institution on the planet. The people who figure that out and contribute to Bitcoin Core and figure out what can be built on top of it is really exciting. Work on something.

We're streaming live.

Steven from Rivetz (maybe?). This is bringing a new capability to the internet and it is the ability to send a secure instruction. It's not authentication and then we open a pipe and stuff some data. What blockchain is enabling is a micropublishing of a financial transaction, but if you step back and see, it's secure instructions in a coreless network.... The device plays an important role in this ecosystem. We don't talk about the device. It's not the abstraction of personal identity on an unknown device is the cyber security problem, the device identity is an important part... the blockchain as a registration authority for devices..  ((What? I suspect people just aren't familiar with cryptography, so that's where this is coming from, even if it has little to do with the blockchain.)) We deal with that because we have fraudsters trying to steal money from all the time. As a reference to other databases the blockchain should be leveraged for those sorts of things.

Q: I am a marketer. What is your plan to push it out and to get those early adopters from the majority to do so? For the most part, people don't even understand the concept of Bitcoin and cryptocurrency. Any thought sof that would be helpful.

A: I could talk from a Circle-specific perspective. We have tried to make things easier and take a lot of friction out for people to get Bitcoin and take away some of the time barriers, you don't have to wait a week to get your first BTC. We have been focusing on making it easier and have less friction. It's still early adopters that are attracted to that product. These are people who want it and are excited about it. I think there's this concept that there's a crossing the chasm metaphor for moving into the early majority... Sean often says that often the products that you create for the early adopters, the products that the early majority wants is a different product. Speaking from Circle's perspective, it's just a baby step it's just a first step on this giant thing we want to do. The adoption curve is always presented like.. the little lines should be really really huge, giant chasms. So whatever features we may have introduced to people who are interested in using Bitcoin and want Bitcoin, may have n bearing whatsoever on the next group of people. We use multisig to secure our BTC, and we have insurance and theft of 100% of your deposits, so if you don't want to take control of your private keys... so to the mainstream consumers, they don't really know they want theft insurance.. even the word key is frightening. There's lots of things like that. They might not even realize they have theft insurance for their US dollar deposits at their current banks. So we need to make these issues disappear. We are interested in person to person payments much more than merchants accepting Bitcoin. Person to person payments is still new, college kids use Venmo and things like that, there's only a few million people using those compared to email or messaging or Skype or whatever which have 100s of millions of users. Person to person payments doesn't have that yet, that's a problem space we're interested in solving. We think the international remittance market is a subset of person to person payment market. The cost of payments will be driven towards zero, it's just the cost of moving data unless Gavin really screws osmething up and we have to pay huge fees. So that sort of happens and that means sending messages that changes ledgers entries anywhere on the planet.. remittance is just a feature. We don't think.. before the internet, we thought sending international mail required lots of postage, there's this hub and spoke routing, that's how it worked, and that's how money works today and it's very similar. I think that just goes away and it's a free service which is good. That probably doesn't totally satisfy you. For a lot of mainstream adopters, they will not live their lives on Bitcoin, it kind of implies that we need gateway access to fiat systems and so on, so it's logical that we would need ot... you have all these walled garden servers, you have AOL and Compuserve, and then they could connect and then universities got on board, and then corporations setup that used to be internal only email. So that's basically these batch file store and forward systems for updating ledgers in different countries, we just need to connect those to Bitcoin and it's kind of like connecting old walled gardens into the internet. They are trying to connect old store and forward legacy systems for ledger updates to update to Bitcoin and get more modern more interesting ledger system. This should benefit everyone. As we are starting to solve problems, other people and other companies can solve problems in this space. So we look at it collectively as an ecosystem moving forward. The last thing we would want to do is create yet another closed network that reaches into a few countries, I don't think that accomplishes what we want in our own business.

Q: What do you guys think about the interaction with the corporate interests like Circle and Bitcoin Foundation? So it's about Bitcoin core, and you're giving the second talk. So how do we manage these power dynamics? How do we make sure we don't see corporate ownership with web browsers in the beginning? How do we push Bitcoin forward in a way that is good for everyone?

A: All of the non-profits that exist around this, there's advocacy orgs like Coincenter, there's emerging self-regulatory organizations, now with the Foundation being focused on Bitcoin Core, there's tremendous... we want to support those, but there needs to be lots of companies contributing and helping. That's how you deal with that I think. You can imagine, like Patrick said, the IETF. Is this an IETF thing for Bitcoin? That's a highly political organization with industry stakeholders. It takes political muscle to get things done there. And Microsoft and Google and others are playing things out there. I think that's a shame. I think meritocracies are good. You see this in other open source projects that have taken scale and taken hold on the Internet. Hopefully it's subject to meritocracy... What really matters is the work that happens. Whatever the process is, whether it's run by the Foundation or whatever, that the process allows for the right code to get in, regardless of what the source is regardless of getting shouted out. Ultimately code will win out.

Cool, thank you everyone.