This series highlights the key insights and lessons from our Digital Leadership series of podcasts. We spotlight the important takeaways from our interviews in an accessible format. The following insights come from Leemon Baird, Founder and Chief Scientist at Hedera Hashgraph, and Founder & CTO of Swirlds Inc. Stay tuned for the full podcast interview with Leemon Baird, in the meantime, take a look at our full library of podcasts.
We’d love to get a little bit of context, and understand a bit more of your history, and what really brought you to your current role.
Well, I’m a computer scientist, I just love playing with computers and algorithms, and that sort of thing. I was professor, I’ve been a scientist in a lot of labs, my PhD is from Carnegie Mellon in Computer Science, and I started a bunch of companies, and all of them have dealt with computer scient projects and algorithms, and interesting things like that. I love solving puzzles, and so I do that for fun, I teach people things about it, and I start companies that use the solutions to the puzzles.
Hedera has garnered quite a lot of attention, and the Hashgraph consensus algorithm of course is the central technology around Hedera. Could you talk about the background of the company?
Originally it was just an interesting problem; could we have some kind of a system for allowing people to share a space on the internet, that they wouldn’t have to pay some server to be holding it, it would just be their own computers, but it would be trustworthy, and it would be super fast. It was an interesting math problem and I kept working on it, convincing myself no, it’s not actually possible, you could be secure or fast, you can’t be both. I kept putting it aside, but it kept coming back and haunting me, gnawing at me, and I just kept working on it and eventually realized If you just store a couple of hashes with each message then you can get a complete history of the messages, and then you can just know the answer without having to talk to each other about what order to put them in. You get the ABFT (Asynchronous Byzantine Fault Tolerance), the strongest security possible, and it is extremely fast, in some senses it’s about as fast as the internet would allow, and the laws of mathematics would allow.
We started a company, Swirlds, to make this into private ledgers, if you want to stand up your own computers and create a private ledger with Hashgraph. We started with that to show that it works well before people will trust it for a public ledger. In 2016 we said, ‘Okay, we are releasing into the world this idea that we had this private ledger, but we need the traction. What we need to do is create a public ledger, that’s really what you want to do, and we’ve got to build trust. So, we’ve got to build a counsel of the most trusted companies in the world and make sure they are diverse, make sure they are in different continents, in different areas, different industries, and that they will even be diverse in time, put term limits on them’. And that was the genesis of Hedera, let’s take this great technology that we’ve been using for private networks, let’s make it a public network, and let’s make this public ledger be run by people you can trust, who are diverse enough that you couldn’t bribe them all to do something bad.
Could you talk about the structure of the governing counsel for Hedera, how it’s different and how it works?
You could just say, well, we’ll just let what ever programmers are most into this, or, let some of the miners that are into this vote, and let them run it. But then of course they get into disagreements, and what do they do whenever they disagree? Well, they split, they have forks, and when your state is splitting that can be a really bad thing.
If you want to digitize the Mona Lisa, you can’t have two different sets of shares in it, you have to know which one is the real one. So, it’s important to actually have a governance system, a system where it will be hard to bribe some of them to do something wrong. So, we said we’re going to have 39 of the biggest companies in the world, Fortune 100 equivalents, and we will have them be the governors, we’ll have term limits – so a three year term, you can only do two terms in a row, so after six-years you have to leave, and the rest get to vote on who will replace you.
We have to have them in different industries, we have 18 different market sectors, and we are saying we’re going to split it evenly across those, so you can only have a few in each one, about three in each market sector, and then no more. Everything we’re doing is about trust, to make sure it is widespread decentralized, in that sense.
We’ll have 39, we’ve announced the first five, we have the largest bank in Japan, Nomura, we have the largest telecoms in Europe, Deutsche Telekom, we have Swisscom Blockchain for technology, and then we have one of the largest online and brick & mortar retailers in Latin America, Magalu – Magazine Louiza, and we have DLA Piper, one of the world’s largest law firms. The whole goal from the beginning has been to build a ledger that is going to last for 100 years; 100 years from now this should be continuing, and so trying to build the trust from the very beginning with this council.
How does the Hashgraph protocol works?
It is fundamentally different from the other algorithms that you’re familiar with, it’s not proof of work, it’s proof of stake, which means we don’t have to buy super computer and waste lots of electricity to do it. The way that it works is, you spread messages with gossip, which is the simplest possible way, you just call people at random and tell them everything you know, that they don’t know. And they tell you everything that they know, that you don’t know, and just do that randomly with people over and over again, so messages spread out extremely fast. This is basically the fastest most resilient way we know to have messages spread out, and if you want to any kind of a ledger where everybody learns all the transactions, you’re going to have to do that, so everybody doesn’t gossip.
What we do is, every time you send someone a message you just tag on two little hashes, which are just compressed versions of fingerprints of the last message you created, and the last message you received. When you send me a message you will tag onto the bottom the hash of the last message you created, and the last message you received from someone else. The thing that we are gossiping is this big pile of messages, but the messages because of those hashes are themselves telling us how we talk to each other. In fact, if I look at this big pile of messages, and look at how the hashes connect them, they form a graph which shows me a history of how we all talk to each other.
When you receive a message, after you receive a few more, you will suddenly know what order that first message was in the consensus order. And its asynchronous byzantine fault tolerant, which means that even if almost a third of the computers are malicious, we’re still fine, we will still make progress, we will come to a conclusion in a few seconds with true finality.
What are some of the use cases that you are particularly excited about?
We’ve had a lot of interesting things like asset tokenization, where you take stock or you take real estate and tokenize it. We’ve had a lot of interest in people looking at track and trace, provenance, so if you’re buying diamonds, are they blood diamonds, did they come from conflict or are they legitimate? Can you buy them in good conscience? Or, if you’re buying food or pharmaceuticals, where was it before it came to your table? Is it really something you really want to eat, or is its background shaky? Or, was it produced by slave labor or oppressed people? You want to know its history, and so you can do the provenance on that.
We’ve had people very interested in credential management, we have revocation service woven through all of our services which allows you to store credentials, and revoke them when you want, and whoever issues you the credential can revoke it, so your driver’s licence, or a diploma, or certificate that says you are a doctor or whatever. We have people using this for healthcare systems, trying to get the ability for any doctor to be able to write prescriptions to any pharmacist, and you can match them up, and have them match up with the patients, so the patients know the credentials of the doctor. If the credentials are revoked, everybody finds out about it. So, these are all interesting things.
Please share the details of the new service, and what can it do.
We have three services, we have crypto currency, and we have files, and we have smart contracts; so, you can run smart contracts on it, you can store files and delete them, and read them. And you can have a cryptocurrency, the h-Bar. Standard things that you would do on a ledger, but they worked really well because they were on this really powerful ABFT consensus engine.
Now, we’re offering a fourth service which exposes the underlying engine. It allows developers to use this new consensus service to basically use our consensus engine, and to use the trust they have in our governing council and our nodes to amplify the trust of some other network that they built. So, this is different from a typical layer, two-things it is different from, a side ledger or side chain, and it is different from sharding, it’s even different from our mirror net, it’s different from all these things. Here’s what happens, you and other people set up your own set of computers, maybe you’re setting up your own stock market, or your own auctioning thing, just for your own personal use. Or maybe you’re setting up something like an auction site for the whole world to use, so it could be public or it could be private, either way.
You have a bunch of computers, maybe you’re playing a game and you have a bunch of computers that are involved in this network playing the game together. The processing is happening on this network, so it doesn’t have to slow down the main network, and the storage of the information is in this local network, for the long-term, so you don’t have to use up lots of hard drive space on expensive main net nodes. But the transactions you’re doing, you send them as messages to Hedera, Hedera puts them in order, puts timestamps on them and sends them back to you along with cryptographic proofs that it’s telling the truth, you can prove to other people that this is the true order. At that point your local network is extremely fast, the throughput of course is going to be enormous because it’s flowing through Hedera, it’s going to be very cheap because Hedera can handle so many per second, and Hedera’s not having to do any of the processing, Hedera’s not having to do the storage.
You have the guarantee that they’re in a fair order that’s useful for something like sales, or auction sites, or stock market, or a market for parts in a supply chain management system. You want fairness, it ensures fairness, and even if the computers in your application network aren’t trusted, if even one of them is trusted then you can trust the whole results, because you have these proofs coming from Hedera. So, really you’re inheriting the trust of Hedera, if you trust that Hedera has more than 2/3rds of its nodes being honest, then in your application network you only need to trust one node as being honest, and if you’re one of the nodes you trust yourself and you’re done. You have 100 percent trust. If there’s ever a dispute you can just go to the court and say, ‘Look, here’s the list of the messages, here’s what we calculated from them, and here’s the proof from Hedera that this really is truly the right ordering for these vestitures', and you can prove it to a court, you can prove it to an auditor.
Momenta Partners encompasses leading Strategic Advisory, Talent, and Investment practices. We’re the guiding hand behind leading industrials’ IoT strategies, over 200+ IoT leadership placements, and 25+ young IoT disruptors. Schedule a free consultation to learn more about our Connected Industry practice.