Factom Q&A: Secure Record-Keeping a Chip Off the Old Blockchain
Yet there are other technologies today that might be just as perplexing to laypeople: take blockchains and cryptocurrencies, for example.
Factom, based in Austin, TX, deals in both. But its mission is relatively simple: It helps businesses, governments, and people securely manage data and keep records.
The value of using Factom’s system, rather than some other data management service, is based in how the company records and publishes data using blockchains, says Peter Kirby, the company’s president. Factom is collaborating with a variety of entities, from healthcare organizations such as HealthNautica to governments and financial institutions.
The term “blockchain” is a metaphor for what the technology does. Factom stores data for its clients using encryptions that are called hashes (more on that later) in blocks, which are then stacked like bricks in a wall or links in a chain. Those chains are made public; they are published on websites, such as www.blockchain.info. Because the data is encoded, no one except the user who created the encryption can decipher it. The blockchain, which was developed for tracking transactions in the cryptocurrency Bitcoin, proves publicly that the record exists, Kirby says.
Kirby recently discussed the technical details of blockchain technology and how his company operates, in an in-depth interview with Xconomy.
Xconomy: What do you do?
Peter Kirby: Factom is a publishing mechanism. It says, I had this data at this point in time. I’m going to publish it to this data layer—this broad distributed ledger that anyone can write to. Not only can I prove at any moment in time that I had that data, I can start talking about it as a process. A bank document with a social security number, and the amount you earned, maybe that gets updated once a month. You can start showing that on day one it looked like this, on day 30 it looks like this.
X: Why is that valuable?
PK: The use case I come back to is this idea that you’re going to build tamper-proof records. Every bank, every insurance company, every government, every whatever: They’re all a big stack of records. Records in a database are only as secure as you can make the database. We’ve seen some ugly hacks into databases: Sony Pictures, BlueCross BlueShield. If your records are only as secure as the central access point, it gets really dangerous when people have access to that that shouldn’t.
X: What’s an example?
PK: We were down in Central America, and we were talking to a country that had a horrific problem where they built a big database for land title records. Some bureaucrats got in there and started giving themselves beachfront property. Imagine if someone could take your home by just changing the name on the legal description. In that situation, if you had a blockchain backend, then basically you have a permanent audit layer. If I misspelled your name, I could go back and change your name. Now I have proof of the misspelled one and the spelled-correctly one. Then, if you sell it to a family member, we’re basically changing the first name. Now I have proof, and I can show the process of any record over time. Tamper-proof is a really big deal for records. That would solve bank transaction problems. That would solve insurance fraud sequence if I know that this particular vender has committed multiple insurance frauds, I can look at every record they’ve done. It’s hard to make record keeping a really sexy thing.
X: Unless you’re an auditor.
PK: Have you met auditors? I wouldn’t describe them as sexy. It turns out, core to every business process is knowing what happened and knowing it hasn’t changed. When we talk about data for blockchains, that’s one of the core ideas: you can do really tamper-proof stuff.
X: What other value does it have?
PK: Another core idea is the balance between privacy and transparency. Say I have a title record I want to make totally public, but your personal financial information, I want to keep totally private. Somewhere in between is a proof of ownership where I want to have your name and a legal description, but not where your home address is or what your company does. That’s something you can do on blockchains in ways you’ve never been able to before. Hashes can keep private data private, yet I can go do an audit for all the title records from a government, and say, look we can see what’s happening to all the titles for a government in a transparent way, without giving away private information.
X: I can’t really wrap my mind around blockchain. It’s not like something tangible, right? What is it?
PK: When we say blockchain, what we literally mean is block of transactions, chained together. What a blockchain does is create a permanent public record. The last thing I want to do is put my social security number on a blockchain. The way you balance privacy and transparency is this idea of hashing. You take data and create a digital fingerprint for it. Then, that digital fingerprint proves that data exists and existed in that moment in time in the same way a cop can dust a fingerprint on my desk and prove it was there. What it doesn’t do is allow you to take that fingerprint and recreate a person from it. That allows you to keep private data private while still proving you were there, or ownership of a document.
X: So that fingerprint—the hash—is just code, like string of numbers and letters?
PK: Yeah, it’s just 1’s and 0’s essentially. Most of the time it’s a 32-bit string of random numbers. It’s vastly improbable—you’d need all the supercomputers in the world working for the life of the universe—to recreate it, just like it’s impossible to take a fingerprint and recreate the person. Let’s say you take a page of text with your social security, your income—a bank document. You put it into a hashing engine—an encoder—and you end up with a long string of digits that represent that document. If I put it into the same encoding mechanism with the same inputs with dates and what bank it’s from or whatever, it would always, always, always recreate those same digits. If the input is the same, the output is the same. But you could never recreate the original. It’s a one-way coding.
X: The blockchain still seems like this amorphous thing.
PK: If you say we publish a newspaper, it’s not like you mean the newspaper is only the platform we stamp stuff on top of. It means that every day the newspaper comes out and it creates the history of the newspaper. The blockchain is just these whole series of transactions happening moment by moment. Together, they form the history of the newspaper—they form the blockchain.
X: So how do I see the blockchain?
PK: At Blockchain.info, I can pull up any record. It’s my real-time blockchain of transactions.
X: What do blockchains have to do with Bitcoin?
PK: Bitcoin is a decentralized network. There are all these people mining Bitcoins. Every time a transaction happens, all these miners agree it gets published to a block. Miners run a piece of mining open-source software, which runs the Bitcoin protocol. That protocol tells these miners to input a lot of computational power, and agree to publish the blocks. You download [the software]. It runs by machine. The machine just cranks at these computational problems all day long creating hashes. Those many, many billions of transactions, or hashes, get stacked up in bricks.
X: And Factom will just be using that?
PK: In addition to the Bitcoin blockchain being a wonderful way to publish financial transactions, it turns out it’s a really wonderful way to publish a lot of stuff.
X: So that’s what Factom is doing, creating hashes out of client data and publishing on the Bitcoin blockchain?
PK: We built it differently, but it’s the same concept.
X: You’ll have miners who publish the transactions?
PK: We have 32 federated servers. Those servers agree on everything that gets published, and for doing that they get paid out in Factoids, our nickname for the tokens that get paid out.
X: Wait, so that’s basically how Bitcoin works, right?
PK: In Bitcoin, in order to get minors to do the hard work, there is a Bitcoin reward. It’s 25 Bitcoins every 10 minutes. It’s not paid by someone writing checks. It’s hardwired in the system that every 10 minutes it pays 25 bitcoins. (Eventually, that reward reduces until 21 million Bitcoins have been issued.) There are Factom servers that are equivalent to Bitcoin miners. The Factom system pays the servers in Factoids.
X: Factom has people running servers who publish the client data for a reward?
PK: Anyone who wants to run a server can. The servers get the Factoids. We’re selling some to the crowd now, we’ve kept some ourselves. We’ve sold about 3.7 million of these things so far. Unlike Bitcoin, we didn’t do a half-life. Every year, the amount we sell increases by a percentage, though that inflation rate decreases over time.
X: How do you get to become a server?
PK: The people who run the fastest, most honest servers get voted in. The voting is done by the users of Factom—the banks, the title companies, and everyone who is using the system. They say, David is running a really clean server, let’s vote for him.
X: How does voting work? Do some clients get more votes than others?
PK: Every time someone publishes data in Factom’s system, they have to spend a tiny thing called an entry credit. One entry credit equals one entry into the system. One thousand entry credits equals 1,000 entries into the system. The entry credits you bought get you the vote.
X: OK, so how do you buy an entry credit?
PK: You can buy entry credits from anyone who has a Factoid. Let’s say there’s 1 million Factoids in the world. Let’s say someone wants to buy 1,000 entry credits, so they cash in a Factoid for 1,000 entry credits. That Factoid gets removed so there are 999,999 factoids remaining. The more entry credits get used, the fewer Factoids there are. Factoids can then go up in value, and down in value. Because they’re getting used up over time, there’s a direct relationship between the value of entry credits and the number of Factoids.
X: So both the Factoids and the entry credits are like a currency?
PK: What is really cool is an entry credit is nontransferable. It’s purely a software license. I can’t sell it to you, I can’t trade it to you. Once you have an entry credit, you’re stuck with them. You never have to touch Bitcoins, you never have to touch Factoids, you never have to touch cryptocurrency. It turns out that is a big deal.
PK: Every lawyer in the world is saying do not touch those currencies—Bitcoins or Dogecoins. The big institutions don’t want to hold Bitcoins.
X: Haven’t some financial institutions invested in Bitcoin-related companies, such as Circle or Coinbase?
PK: True. BBVA and USAA invested in Coinbase and Goldman Sachs invested in Circle, but these were equity investments in Bitcoin companies, not the coins. These Bitcoin companies are standard C-Corporations and pretty vanilla investment vehicles. As far as I know there are no major financial institutions that have purchased a large amount of Bitcoins or any other cryptocurrency.
X: Why, because the price volatility could make you lose money?
PK: You could lose money, you could get sued, the government could decide that you’re doing money laundering—it’s a money transmission without a license. This blockchain stuff is cool tech. This is one of the limitations of adoption: They’re looking at this world and saying, stay away. What the entry credit does is says this is purely a software license.
X: So how do you make money then? If I have a bunch of Factoids and sell them, do you take a commission?
PK: Factoids convert into entry credits cleanly. However, we’ll create a store front and we’ll probably take a small commission on those. That storefront also lets us sell other tools on top of Factom. I imagine that there’s a lot of tools that a title records company will need that a financial app will need also.
X: What kind of tools?
PK: A really simple one is identity management. How do I know that you are you? My title record might want to know who customers are, who an attorney is. Identity management is a really simple tool that I can build on top of Factom and sell to other people.
X: So you sell a subscription?
PK: Yeah, we get to be blockchain as a service. We sell them tools and consulting. I get to sell those tools at the value they create. It’s not cost plus sales. It’s that I’m solving this pain point for a bank. How much value is there in that?
X: The more tools that are created, the better it is for you, it seems.
PK: Factom thrives when 1,000 businesses get built on top of it. For the guy who holds Factoids, the more cool stuff that gets built on top of it, the more entry credits get sold, the more your token tracks to that value. For the guy who is building the business, he’s got a great customer and he understands that subject and he gets to collect money. For us, we say, in order to do that, we’re going to sell you this set of tools and it’s going to make it easier to do what you want. What I love about the model is that it gets everyone’s incentives lined up. That’s the ethos of open source: more smart people working on a project makes it better.
X: You said you’ve sold 3.7 million Factoids already?
PK: We had some people pre-buy these tokens from us. The Decentralized Applications [Dapps] fund is the first VC fund to be exclusively token-based. They did a syndication round and that raised us the initial funds to get us live. We’re selling tokens to the crowd right now. We’ve sold about 3.7 million of them, which translates to about half a million dollars. The Dapps fund and syndication partners bought about $750,000 of tokens. The tokens they receive are determined later. That’s all for the foundation, the open-source technology behind Factom.
X: Are you raising more for building the subscription service?
PK: For the company that’s built on top of it, we’re doing a seed round right now.
X: Are you finding much interest in blockchain?
PK: Every major institution I’ve talked to—the big consulting groups, big banks, big insurance companies, governments—most have a blockchain working group, trying to figure out what to do here. When I see that, it reminds me of 1994: Every company had an Internet working group. It tells me two things: 1. It matters. 2. It’s coming.