Epicenter - Learn about Crypto, Blockchain, Ethereum, Bitcoin and Distributed Technologies - Paul Brody: Internet of Things and the Democracy of Devices

Episode Date: April 20, 2015

As the cost of computing continues to plummet, the number of connected devices is increasing dramatically. In the future powerful computing capacity will be in billions of devices ranging from the lightbulb to the car door to new wearables. This new internet of things (IoT) will dramatically change our lives. Paul Brody, a Technology Strategist and the former VP of Mobile and IoT at IBM, joined us for a discussion of where the IoT is going. We discussed their white paper on why a decentralized, blockchain-based internet of things could solve the problems of the current IoT in terms of privacy, scalability, and business models. Topics covered in this episode: What the Internet of Things and its core value proposition is Why the current IoT business models are broken How a decentralized IoT infrastructure would be more secure, scalable and economically feasible What role the blockchain could play in a decentralized IoT The three components of IBM’s ADEPT proof-of-concept Episode links: IBM Democracy of Devices Whitepaper Paul Brody IFA Talk IBM Empowering the Edge Report IBM Empowering the Edge Use Cases IBM & Samsung ADEPT Demo at CES This episode is hosted by Brian Fabian Crain and Sébastien Couture. Show notes and listening options: epicenter.tv/075

This episode of Epicenter Bitcoin is brought to you by Shapeshift.I.O. With no account or sign up required, it's the easiest way to buy and sell light coin, doche coin, dark coin, and other leading cryptocurrencies. Go to shapeshift.io to instantly convert all coins and to discover the future of cryptocurrency exchanges. Hello, welcome to Epicenter Bitcoin, the show which talks about the technologies, projects, and startups driving decentralization and the global cryptocurrency revolution. My name is Sebastian Kuchua. and my name is Brian Fabin Frient Feren.

Today we're going to talk about the topic that may seem to some of our listeners as a little bit off topic, which is the Internet of Things, but there is a very interesting connection here in that some people, including our guest today, think that the blockchain is going to play a really important part in that.

And we have, I think, perhaps the best person on this topic today on the show, who is Paul Brody, so thanks so much for joining us. Thanks for inviting me. So Paul was, until recently, he was the vice president of technology, no, of mobile and Internet of Things at IBM. And IBM wrote a very interesting white paper on how the blockchain will play a really important or could play a really important part in the Internet of Things and sort of saving the Internet of things

of the future. And they've also done an interesting demo at Adapt and Develop a Protector. type, which I think is Ethereum-based. So we're going to have a chance to dive into that today and getting an understanding of why these two technologies should connect and could work together so well. So Paul, can you, a lot of people will probably not have a very good understanding of what the Internet thinks even is. Can you give a brief introduction to that topic?

Sure. So the internet of things, the idea behind the internet of things is that we've been, over the last 20, 30 years, computing prices have been coming down and down and down. And we've gone from mainframes to mini computers to servers to PCs and phones. And now we're really starting to talk about connecting up all of the devices. We can start to put computing power in everything. Door knobs, light bulbs, desk chairs, you name it. If it's a thing, it's going to become smart and connect. And so the Internet of Things is really talking about this era where we're right on the edge of now,

which is the one where every device, everything that's around us, is going to be smart and connected. Is that desirable, in your opinion, that everything around us is connected? Well, you know, this is actually a really fair question. And one of the most important things that we spend a lot of time thinking about is, is it desirable? Why do we want to do that? And I think very simply the answer is it's only desirable. if it makes the thing in question better. If it's not making it better,

then it's not necessarily a great idea to connect it. That being said, I'm a technology optimist. I really think done correctly, almost everything can be made better if we make it smarter and connected. Yeah, I mean, I think the Internet thinks there are also these visions know of some sort of totalitarian surveillance

or robots taking over. I think those are really closely connected also with that. I remember there are a few science fiction books. Perhaps you know them as well from a guy named William Hurtling. I like quite a lot because he tries to be sort of very realistic in terms of how it could happen. And of course there, the Internet of Things in a way it starts to play a really important role

because then AI can basically use the Internet of things to actually start having, you know, basically hiring their own data centers and started building their own things and getting information. So it's, they're definitely also these crazy futuristic scenarios. A lot of them are very closely related to the Internet things as well, no? Well, I, for one, don't think they're all that crazy. And if anything, what we've learned over the last couple of years is that we live in a surveillance state, which is more extensive and frankly, at least for me, more frightening than anything

that could have been imagined in the novel 1984. The level of sort of detailed surveillance on us is really astonishing. And if there's one thing that I personally am concerned about, and I think a lot of my colleagues at IBM wanted to see, was we want the promise of the Internet of Things to be realized without creating a kind of dystopian nightmare of surveillance. People should be able to have and enjoy smart connected devices without fearing for their privacy or security.

So do you think this is going to be sort of an upstream battle because I presume there will be some resistance from institutions like NSA and the US government when it comes to trying to keep the internet things decentralized? Yeah, probably to some extent there will be a lot of concern about this, but I think in the long run, people who really understand security, understand that you cannot selectively weaken security. Security is the same for everyone. If you weaken security for the purpose of a government agency, even if it is with the best of

intentions, you effectively weaken security for everybody. And so I think on balance, the industry and many reasonable people are moving towards the view that good security, it's good for everyone. Beyond just good security, if it makes more sense, and we'll talk about the contents of the white paper, but if it makes more sense for devices to just talk to each other

rather than going through some centralized server, then it makes more sense for the technology to be designed that way, regardless of any security implications or any sort of like surveillance implications that there may be. Absolutely. We are,

the convergence of a number of really big, both technological and social trends, which are driving us towards a new era in decentralized systems. And of course, there's the question of trust, and frankly speaking, I think we live in an era kind of almost after trust. The internet started as an environment where academic institutions and government agencies operated in a very high environment of trust. We now live in an internet era. where I'm not sure I can trust any of the participants on the internet.

So how do you engage in computing after trust or what we sometimes call trustless computing? The second, though, is how do we architect systems in a way that's cost-effective and sustainable? And this is one of the things that really led us to thinking about the blockchain as a mechanism for managing the Internet of Things, because we think it has really enormous, dramatic cost advantages over traditional. centralized infrastructure. I mean, I think that's the part of your WIPA that makes me the most optimistic, because I think I can certainly agree with you that, you know, this surveillance state

that we are in and moving increasingly, or it's getting worse, that is a, you know, it's an extreme threat to liberty and democracy, etc. But I think if we have the sort of economics as well on that side, right, if there is a strong economic incentive towards decentralization, then I think that will just be something that will not be stoppable, right? I mean, if for companies it makes economic so much more sense that it decentralized, then even the NSA and the US government won't be able to keep it centralized. I think that is a big clear in my mind. If that's not the case, right, if maybe you have a certain tradeoff where you prefer from a security privacy perspective, you have it decentralized, but from maybe an economic

perspective or, let's put it, security perspective from the U.S. government side, you know, they prefer it centralized. Then it becomes much more difficult. But I think this made me very hopeful about your paper. So I think the economic case for decentralization is really powerful. One of the things that got us interested in this is we've been thinking about, and I should be clear when I say this was reflecting my time at IBM. In our interview today, I'm only going to talk about my personal perspective, I'm definitely not representing any official viewpoint of IBM, just my personal opinion. But in my time working there, we started to hear from clients who were talking about how much it was costing them to maintain centralized infrastructure for connected devices.

And we started to think about whether or not that was necessary. In the early days of creating connected devices, we looked at the cost of computing power on the device, and we looked at the cost of computing power in the centralized environment. and we deliberately architected systems to move as much computing activity as possible into the cloud. At the time, that was the right thing to do. Computing power was more expensive, and in the cloud you could utilize it more efficiently. That incentive to centralized infrastructure is now flipped around. And what's flipping it around is the end of the embedded chip.

So historically, if I was going to make a smart device, I would have done so on an embedded chip. Today, it's rapidly becoming cheaper to put an entire system on chip, basically almost an iPhone or an Android phone on every device. And with that much power on every device, it's simpler to do the development in software, to do the customization in software, and to put kind of the entire cloud infrastructure onto the device.

So the economics of decentralization have changed radically over the last few years just as a result of Moore's Law. So that's basically because, yeah, we had Moore's Law when it comes to chips and computing power, et cetera, but there hasn't been something comparable when it came to data and the ability, like the, especially the cost of transferring data. Is that the reason? So what's really driving it is the cost of two big things, right, on the technology side. One is from a hardware standpoint, historically, the way companies developed embedded chips was a big non-recurring engineering cost, followed by a very low marginal cost for each chip. Now, you know, I can buy a dual core system on chip with LTE, Wi-Fi, and GPS for about $10 a piece in volume.

It's really hard to beat that. So you basically have an entire server that you can put on a light bulb. In fact, I've got some friends at a Bitcoin mining company that actually reprogrammed one of their smart light bulbs to mine Bitcoin. Now, I don't know that that's a great use of a light bulb, but it shows you how much computing power is sitting on even the smallest embedded devices these days. So the hardware capability exists, but what was missing is a trusted, simple software architecture

for managing a complex distributed application. And when we really started to understand Bitcoin, we realized it's a financial system. But if you take out the financial component, it's really an amazing decentralized transaction processing engine. And so we had the hardware and the software together that creates the ingredients to shift from a centralized world to a decentralized one. Those last more to talk about. But before we do that, let's take a short break to talk about Shapeshift. ShapeShift is the fast and easy way to buy and sell alt coins. If you've ever tried to buy and sell all coins, you know how complicated that can be.

You have to find a reputable exchange, create an account there, place an order, wait for the order to be fulfilled. That can take a long time. Not with Shapeshift. Shapeshift allows you to convert about 25 different alt coins, and they're adding new coins all the time. So you just go to their website, Shapeshift.io, and use their currency conversion tool, which looks a lot like Google Translate for cryptocurrencies. You choose the currency you want to convert, so for instance, Dogecoin and also the currency you want to receive. So Bitcoin, for instance. And hit start.

Once you do that, you'll be presented with a Dogecoin address and QR code, which you would send Dogecoin to. And those would be converted and sent to the Bitcoin address with no confirmations needed. So it's really fast. It takes anywhere between 30 seconds to a minute to, for the first. for that transaction to go through. And you don't even need to create account. They don't even want your email address. So your privacy stays protected.

And did I mention it was fast? Because it is fast. It takes no time at all. So ShapeShift is the fast and easy way to buy and sell all coins. And we'd like to thank them for the support. Go to ShapeShift.io to give it a try. So before we kind of dive into the blockchain role, perhaps, can we spend a little bit more time on the idea that there are specific threats to IOT?

I mean, it sounds almost like when you read your paper that if it keeps continuing on the path we're wrong, you know, this is not going to end well and it's not going to be financable. So what are those threats? So the Internet of Things is a great idea, but as you pointed out, there's a lot of reasons why it might not get as far as we'd like. One of them is fear about security and privacy. We talked a little bit about that. But almost as compelling as the fears about security and privacy are very legitimate concerns about the business model around the Internet of Things.

A lot of folks have this idea that they're going to sell a device. And on top of the device, people are going to pay a subscription. And then maybe to add to that, they're going to have a revenue stream from either advertising or selling user data. And as we started to look at what's actually happening in the marketplace, we realized that these business model concepts are really, really optimistic, probably way too optimistic in a bunch of different ways. First of all, consumers and enterprise customers don't seem to have a lot of interest in never-ending subscription fees. So the devices that do well in the market are the ones that often come with no ongoing subscription fees. So that's one hole in your business model. Secondly, the business of selling advertising data or user data is pretty risky because in the internet, the economics of information are different than the economics of physical goods.

Physical goods have a marginal cost that's not zero. But if I've got two different providers who are selling roughly the same information, it could be how many parking? spaces are near me here in San Francisco or where are people walking around and shopping, if I can get two or three different providers of that data, the marginal cost is zero and the market clearing price goes to zero. And sure enough, when we talk to people who are in this business, over and over again, what they said was we're not getting the subscription revenue wants and we're not getting the advertising or the marketing data revenue we want. And so revenues much lower than expected, cost, on the other hand, much higher than expected because it costs a lot of money to

maintain a centralized data center. So you put high costs together with low revenues and you get a money losing proposition and that's scaring off a lot of companies from making more investment. So in that case, what essentially the proposition you're making is that the business models associated to sort of the service side of the internet of things is not, um, is not scalable and it doesn't work long term. That's right. It doesn't. And the experience of mobile phones and computers is a little bit misleading because if you take a mobile phone, it comes with a lot of free cloud services.

But many of these phones, they last for one or two years. So if your cloud service costs $2 or $3 per user per year, it's only a few bucks out of perhaps a retail price of $400 or $500 or more. That's a drop in the bucket. But when you're talking about the Internet of Things, well, I'm not going to replace my light bulbs and my doorknobs every two years. An LED light bulb has an expected lifespan of about 15 to 20 years in a residential environment. And it's selling for somewhere between $15 and $30 today. There's no margin in that product to subsidize 30 years or 20 years of cloud services.

So the business model, at least as many people can see if it is broken. I was thinking about this, right? So perhaps there is not enough margin to sustain the hosting of cloud services and having that service provided to the users over that long a period of time. However, I'm not sure I agree with the fact that these devices aren't going to get replaced anyway because people like to have new gadgets and new devices and new toys or whatnot. So that, like this is probably a little example, but this light bulb, although it has a lifespan of 20 years, when Phillips or Samsung or whoever comes up with the next version that, I don't know, has disco ball functionality or strobe light functionality and that's desirable, people are going to want to update to that. And also, you know, people are adding new light bulbs because houses are getting built because they may have new light fixtures in their homes and things like that.

So it's not as though these companies are not going to keep selling these products. They are going to keep selling those products, but the installed base is going to keep going up. And in fact, we did a little bit of research. The average age of a car on the road in the United States is 10 years. The average age of a house in the U.S. is currently about 40 years. And by the way, that's a lot younger than the average age of a house in Europe. So it's true. Installed base keeps going up, new product keeps selling.

But the fact that new product is out there does. doesn't eliminate the fact that you have to support the product that is still in the marketplace, that's still being used. And so what we found was that, you know, over time, this infrastructure adds up to potentially a crushing financial burden, hundreds of millions of dollars for large enterprises that might be supporting potentially billions of connected devices. And so one of the things that we said about doing was starting to think, what can we do to drive cost down?

Because we don't see a lot of appetite for spending a lot more money on these services. Surely, you don't want to... I mean, I love a disco ball. I'd like a LED disco ball in my bathroom. That would be great. But do you want to pay an annual subscription for it? I don't know if I pay a subscription.

Maybe a small subscription. Yeah. I mean, even if you pay a subscription for the light bulb, right, are you going to pay a subscription or want that fancy new thing for the light bulb and the washing machine and the door knob and, you know, the bedding sheet and the shower and... That's right, because these subscriptions just keep piling up and piling up and then all of a sudden you're in a situation where you're paying. And also that novelty thing.

I mean, of course, in some instances that will work, right, and for some things. But you can't expect that overall, for example, product replacement cycles and that sort of stuff will dramatically shorten just because, you know, it's smart now and can do, has like a new feature, something like that. At the end of the day, you know, I have a smart home because I'm, I love to play around with this stuff. I can switch on all the lights in my house. I even set up a script so that when I pulled up to the house, all the lights would come on and the door would unlock. And that was really cool the first time I did it. And then one night, I came home after everyone else had gone to bed and that switched on all the lights in the house and made me really unpopular

and that script was switched off and I haven't really gone back to figuring out what to do. But, you know, at the end of the day, it's nice to have smart functionality, but in many cases, it's still just easier to get up and switch on the light. And so one of the other things that we really had to think a lot about was why? Why do we want this and what is the value proposition around the Internet of things? I'm sure you could probably have some sort of IFTTT script that figures out if people in your house are also sleeping and if they are that it wouldn't turn the lights on. Well, I'm the, I'm the biggest I have TTTTT fanboy there is, but I need if then else and or

and it doesn't seem to offer quite all of those things yet. If, say, for instance, Smart Things or Nest is not providing the great UI and all the services bundled, cloud services bundled with their smart filter monitor or whatever, because they don't want to take on the cost of providing that service for the life of the product or what have you. Then who are we proposing that they're just providing the hardware that has, I guess, APIs or whatever that you can plug into. And then other companies are specializing in providing the right UI to have interaction between your devices and nice dashboards and all that stuff. That's a tough question to answer because I don't think that we know what it looks like yet. We're in an era of great experimentation.

I think we know a couple of things for sure. Number one, nobody is going to live in the Samsung house or the Whirlpool House. We cannot live in a single ecosystem or a Google house. And so anything that ultimately takes hold is going to require a huge amounts of interoperability, which is something that most people haven't done a lot about yet. Secondly, I think people are saying that people are still. still struggling to nail down what the killer application is for smart connected devices. And until people really get their hands around that, there's going to be a lot of fumbling around.

So can we talk a little bit about that, the question of what the core value proposition is of the internet of things? Absolutely. I feel very strongly that I think there's two big ones. first of all when it comes to industrial applications or even consumer applications the most important value proposition around the Internet of Things is asset utilization

making the things that we own do more for us and then when it comes to individuals to our personal things especially a subcategory of Internet of Things which is wearables in my mind it's really all about self-improvement yeah that's actually

an area that I'm personally very sort of interested in involving because I'm also one of the organizers of the Quantifat Health meet up here. And so I've been sort of, you know, closely, following that very closely. And of course, it's their potential is very big there, although it's, I think today, a lot of these devices maybe don't quite deliver yet, you know, what we want them to do. So I definitely see the use case there. Can you elaborate a little bit more on the you know, making our assets, making our devices do more? Absolutely. So right now we live in a world where consumers and enterprises own assets.

We own lots of stuff. The fact is, our stuff doesn't do that much for us. If we think about offices, we think an office is busy when it's in use 35 to 40 hours a week. The week is 160 hours, right? The average car is only in use of. about 9% of the day. And so I believe that asset utilization, getting more utilization out of cars, houses, offices, these are the things that create really colossal value.

If you look at somebody like Uber, part of their value proposition is absolutely great user experience. And part of their value proposition is a little bit of regulatory arbitrage, right? They're skipping out or bypassing some regulations that taxi companies normally have to deal with. But a big chunk of their value proposition is utilization. Uber creates a liquid market for customers and taxi drivers in which to engage. And the data that we can see shows that people who participate in these markets, the drivers, for example, are quite a bit busier than drivers who use either traditional taxi services or traditional limo services. So it's all about increasing the utilization of assets and getting more value from them.

The same thing is happening with Airbnb or VRBO for vacation homes. As the Internet of Things goes forward, they're going to be instrumenting billions of devices, cars, trucks, storage spaces, offices, all of those can be utilized much more efficiently. And of course, the interesting thing, right, is that you start having marketplaces in those things, right? and you can have maybe different pricing, right? I mean, one of the things Uber is known for is that they adjust the prices depending on the demand. And so you can kind of affect both supply and the demand.

And I guess you could imagine something similar of let's say office space where then on the weekend or an evening, it's cheaper per hour. And then maybe some people will shift their working hours to accommodate cheaper supply and those kind of things. So that's, I think that's really interesting. because you start having all these interactions and markets and possibilities that, you know, they're just not financial transactions there.

There's no market. There are no, they're not really used in that way at all today. That's right. It's too difficult today to create a digital marketplace for relatively smaller transactions. If you take something like borrowing somebody's car, if I want to borrow your car for two hours, I've got to meet you. I've got to give me the keys. I've got to get it, bring it back to you, return the keys.

Here in San Francisco, there's a startup called Get Around. They have created a little electronic device that you can put in your car. And from that electronic device, you can locate the car you want to rent from your smartphone. Using your smartphone, you can unlock the car, drive it away. Mileage and payments are all calculated by GPS. And when it's returned, the car is locked and the owner is told exactly where the car was returned. And that's a perfect example of how the Internet of Things makes something that would be too difficult to do today

actually pretty straightforward. Maybe one other question regarding the sort of decentralized vision that you've been painting before maybe we can dive into that more. So you talked about surveillance. Do you think that consumers demand this and will demand this? like a sort of surveillance proof and privacy honoring infrastructure? I think that the average consumer does not spend a lot of time thinking about surveillance or data privacy. However, I also believe that the average consumer, when it comes to buying these kinds of new services,

they turn to people that they trust and respect and they ask them, what would you buy? Do you think this protects my privacy? And I think a small number of people who care passionately about this have come to believe that these highly centralized systems represent a major security risk. And so I think that small group of people is not only putting tremendous attention into this, but they are guiding the market in the right direction. And so I think ultimately the winning strategies are going to be ones that respect security and privacy. Cool. Well, diving into your white paper bit. So the title of the white paper is device democracy.

How did you guys come up with that title? So the origin of the title actually came right at the end of the paper. We spent a lot of time asking what are the business requirements, what are the technological changes going on, and we realized that decentralized systems, especially the blockchain, are the right of, mechanism and tool for creating a massively scalable, secure internet of things. And the blockchain, in a sense, is a kind of democracy.

All the participants in the blockchain contribute to its security and its effectiveness because they do, in Bitcoin they do transaction mining. And we foresee a similar kind of future where the strength of an Internet of Things blockchain or Internet of Things blockchains, there might be many, that comes from the participation of all of those. So we love this idea of thinking about it as device democracy, decentralized and really in the hands of the users and the device owners. Will it be a role for mining in the Internet of Things? Absolutely. It's not going to be the same as the mining that goes on for Bitcoin,

but there is going to be mining. And I think more specifically, there's going to be transactions. Bitcoin is all about transactions. I'm paying you a little bit of money. You're paying me a little bit of money. we're exchanging services. In the Internet of Things, we're going to have lots of contracts and transactions as well. They're not all going to involve money, but they are similar kinds of things.

I'm going to authorize you to borrow my car. I'm going to let you unlock the door to my house. I'm going to connect my smart watch to my smart lock, and every time I unlock the door in my house, that's a transaction. And it's a transaction that's operated between different devices. And so what we realized was the blockchain,

Bitcoin is a financial transaction processing system. The Internet of Things is just a transaction processing system for connecting billions of different devices. And so those two things are actually very similar and complementary. And in Bitcoin, mining serves two purposes. One is to provide security, and the other is to manage the introduction of new Bitcoins into the economy. So in the Internet of Things, we don't necessarily want to restrict the number of new devices. So we're not going to make our mining arbitrarily more difficult, but mining still needs to be done for the purpose of securing the blockchain. So will the devices be doing mining?

I believe they will. I think every individual device will be able to do some mining. There's so much processing power. I mentioned before, I know some friends who got one of their small. smart light bulbs to mine Bitcoin. There's processing power everywhere. There's no reason why it can't be used to help secure the internet of things.

So this is interesting because this is something I brought up with Vitalik. Brian, I don't know if you remember, but ever since I've been interested in Bitcoin, I've had this idea that one day all these internet-connected devices would be mining. And when I asked Vitalik about this, I don't remember exactly in what context, his response was,

well, no, because usually these devices are all embedded systems. And the thing about embedded systems is that they're designed for one thing only, and there's very little extra processing power that's left over after you're done just doing that one thing. And this sort of flips everything around because you're proposing that, in fact, we're not going to be using embedded systems. We're going to be using these systems on a chip, which would have a lot more processing power out of the box. and a potential for, I guess, these empty cycles or cycles not being used, which we could use for mining.

That's exactly it. We are about to experience a complete flipover. Vatallic is exactly right. Almost every embedded chip today doesn't have any power left over to do anything else. But typically, for example, I worked with an appliance company. They spend an average of $5 a chip today, $5.5. per product today on embedded electronics. That is pretty close to the $7 to $10 you're going to be able to get a system on ship for.

Embedded chips take a long time to design and qualify. So within with Moore's Law, within a couple of years, it's not just going to be cheaper to put an entire system on chip on your refrigerator or a blender or your doorknob. It's actually going to be faster because you can do all the customization and development and software. And so almost overnight we're going to go from barely enough processing power to run the refrigerator to a thousand times more processing power than is necessary to run the refrigerator. Okay, but now with regards to mining

and Bitcoin mining specifically, I presume then that you're not proposing that we mine the actual Bitcoin blockchain on these devices, but we're talking about some other blockchain that doesn't require A6. That's right, because Bitcoin contains in it the feature of escalating mining complexity, for the purpose of,

it contains escalating mining complexity for the purpose of limiting the output of new Bitcoins. The requirement today on Bitcoin is far, far beyond what's necessary just to be cryptographically secure. What we just want is sufficient security that the blockchain itself functions not to make it arbitrarily difficult.

So here, of course, the question comes in with the electricity costs, right? because if mining, you know, if the point of mining for one of those devices is basically, is that to generate some revenues in terms of earning some sort of currency? No, the purpose of mining is to claim kind of your share of participation in the blockchain, to secure the blockchain, right, and to make sure that transactions are processed correctly. The mining, what will protect and secure an Internet of Things blockchain from hijacking by any individual participant is if we have billions and billions and billions of devices out there, and those billions and billions of devices are all doing their own mining, then the amount of computing power required for a hostile participant to launch a 51% attack is impossibly large. And so by having every device do some of the transaction processing and some of the mining,

we are helping to secure the overall future of the blockchain for the Internet of things. But then we will still need some sort of proof of work that is at least semi-resistant to A6s, right? That's right. So you need reasonable, you need enough complexity for cryptographic security, but without it being arbitrarily complex. What I would argue is that it doesn't need to be a great, great deal if you're talking about having billions and billions and billions of devices able to do mining. Okay. Now, that's fascinating. I actually remember when I was at first Bitcoin conference,

I had dinner with Manny Roosevelt. He has been doing a lot of work on mining and mining incentives and stuff. And we had this conversation back then as well, you know, will the light bulb be doing mining and stuff? Super fascinating. So, because of course this is for a totally different for a totally different reason, right? Because then the device is not mining to generate revenue, but the device is mining to sort of assure its right to keep participating in this network

and communicate with other devices. Exactly. Okay, cool. So we mentioned transactions. It's unclear to me what exactly these transactions would represent. unless there is some sort of financial transaction going on. But if my light bulb needs to, I mean,

if my smartphone needs to tell my light bulb that I'm walking into the house, there's no financial transaction there. What does that transaction contain? So a transaction, if we step back on with a financial transaction is a transfer of money, but in computing terms, a transaction is just an exchange of information, a single step that's processed.

So a tweet is a transaction. for Twitter, right? A like is a transaction for Facebook. A door unlock is a transaction for an Internet of Things device. So that's one thing, which is that actions are effectively transactions between devices. So that's one way of looking at transactions. The other thing is that although it might not be financial, we believe that many of these devices will actually participate in non-financial marketplaces. So one of the use cases that IBM and Samsung worked on together, is how washing machines and dishwashers and televisions and other home appliances work together to manage their consumption of electricity and how they autonomously connect to each other

and make sure that the house as a whole isn't consuming more electricity than, say, for example, might be being produced by the solar panels on the roof. Okay, so, but those transactions wouldn't be part of a blockchain, right? I mean, unless it's actually sending money around. No, they could be part of a blockchain. We could have an energy blockchain that would allow devices and neighbors to trade electricity with each other.

And in fact, there are a number of startups already that are looking at how to create distributed markets in energy, in bandwidth, in storage, right? You name it, these kind of digitized commodities. We can create both barter-based marketplaces and financial marketplaces around them. And in a world where even your light bulb has as much power as the original iPhone, there's no reason why that's not possible.

So I guess one big challenge as well would be scalability with that, no, because I mean, I think currently the Bitcoin blockchain, right, is very limited. And I mean, of course, people are trying to come up with more scalable blockchains and cryptocurrencies and Bitcoin itself is trying to scale more. But that's very much of an unsolved problem. And transaction times as well, because, I mean, if you need to send a transaction and an action needs to happen immediately, then that transaction needs to be fairly fast. Yeah, great point. So how are these, how is that going to be solved? So I would say we came up with a mixture of things that we definitely could solve and things that we couldn't solve right away.

We talk about, I mean, just to give a bit of an analogy, we talk about Moore's Law all the time. But the fact is, we don't exactly know how the next two or three generations of chips are going to be made. but we know that there's progress. I'm a real optimist. I think the progression that's being made on the blockchain and improving cryptocurrencies and distributed systems is enormous. So I'm confident that some of the basic issues of block size and transaction processing volume can be overcome.

Very specifically in the work that IBM and Samsung did together and it's detailed in the new white paper we just put out, IBM just put out called Empowering the Edge last week. we looked we didn't just use the blockchain we also added a couple of other pieces to address a specific point so we looked at the blockchain for transactions of record we looked at a system called telehash which is a distributed messaging system that enabled quick transactions like i'm walking up to the door and i want to unlock it and we also enabled bit torrent as part of the architecture prototype for transferring very large file volumes So any complete integrated Internet of Things architecture is going to have more than just the blockchain. But I think the foundational underpinning is going to be the blockchain. And of course, one of the cool things is going to be that if you have the blockchain, then of course the LIPAL can have a Bitcoin address or some cryptocurrency address and make payments.

And I guess also, I mean, you talked about sort of the barter use case. but then it could also be that right the light bulb sells spare capacity to someone else earns money back to for its owner I also thought one of the things while I was reading the paper that sort of came to mind that could be kind of interesting right so if you it's sort of a new way of financing you know financing devices financing assets is if you sell them to a consumer, but sort of at a reduced rate, and in return, all the revenues from maybe excess capacity that's being sold on a marketplace goes back to the manufacturer or maybe some third financing company, right? So you could actually imagine that, you know, connected devices would be significantly cheaper because of that. I think so. And I think we'll see a lot of, as the blockchain spreads, one of the things that's great about the blockchain is,

It's an incredibly transparent and very secure marketplace. So when you see a history of transactions in the blockchain, you can trust it because it's been secured by all the participants. And against the trusted set of transactions, it's possible for us to do all kinds of insurance, financial services, right? In general, we talked about at IBM the creation of the economy of things. right so from the internet of things just connected devices to a true economy where all of these devices both are interacting autonomously with people and with each other to create different marketplaces yeah i mean this is sort of touches on very similar to uh what uh mike herne has talked about extensively in talks and on this show as well is is a world where you have it in internet of

things and an economy of things where devices uh are interacting with each other and there's no human intervention and just transacting with each other. Today's magic word is connected, C-O-N-N-E-C-T-E-T-E-D. Head over to Let's TalkBitcoin.com to sign in, enter the magic word, and claim you're part of the listener reward. Now, with regards to Adept, now you mentioned the three components, blockchain technology, telehash, and BitTorrent. Can you go more in depth as to what exactly is adept?

And perhaps it would be interesting for you to lay out one specific use case or perhaps a use case that we can imagine using Adept. Sure. So Adept stands for autonomous distributed peer-to-peer telemetry, I believe. And the idea behind it was to create a proof of concept that using the blockchain can allow us to do many of the same things that we see

in traditional centralized infrastructure systems. So if you think about if this, then that, or smart things, or a lot of the other centralized internet of things services out there, what we wanted to do was create a couple of use cases, not amazing state-of-the-art stuff, but show that using the blockchain, we can deliver the same functionality in an entirely decentralized. way. So we wanted to be able to move big files, we wanted to be able to secure transactions, and we wanted to be able to have kind of quick actions take place between devices. So the blockchain is there for securing the overall system and providing the management of contracts. Telehash is for

rapid transport and quick action, and BitTorrent is for moving very large files asynchronously. And in cooperation with Samsung, we developed a set of use cases, which we showed for it live at CES back in January. And those use cases do very simple things. For example, showing how a washing machine can diagnose its own mechanical fault, check its warranty status with the manufacturing company, and then go into an online marketplace and request repair services. all of those things were done without any centralized infrastructure. Okay.

So that was a good example of how it's not earth-shattering stuff. Our purpose was to show very mundane things can be done in exactly the same capability set as a centralized system, but without requiring any server infrastructure at all. So for each of those actions that you mentioned, in what case is the telehealth, messaging system use, in what cases the blockchain use, and what cases BitTorrent use? Because it's unclear to me when we would simply need to send messages through telehash or when we'd need to actually initiate a transaction on the blockchain. So, for example, in the blockchain, we use it to establish the kind of originating records

of the device and to establish contracts between, say, the device, the device owner, and the device manufacturer for things like... ownership and warranty. Once those things are established and recorded in the blockchain, then if we want to do a very quick command like open up the door or unlock this process or start a washing cycle, that can be done by Telehash provided that it's checking against an already recorded contract on the blockchain. Right, okay. Blockchain is kind of the system of record. Telehash can be used for quick transactions and lookup, And then something like BitTorrent is very useful.

Let's say you want to transfer a huge file of diagnostic data to the manufacturer to figure out what's wrong with your washing machine. You don't want to load that whole file into the blockchain. You just want to move it efficiently across the network. BitTorrent is amazing for that in terms of distributed file transport. Cool. This is so awesome. Yeah. I mean, so in this case, they used Ethereum, I believe, the Samsung demo.

Yeah, the IBM Samsung demo was built on Ethereum, Telehash, and BitTorrent. Cool. So what's the plan with that? Is the idea to actually go from a proof of concept to an actual implementation that's going to be deployed? So the main plan with ADAPT was to get our hands wet and really get our feet wet and really understand what can this work and does it work. And in fact, this last week, IBM published two new white papers on the topic. One's called Empowering the Edge, which is all about kind of the big picture of the lessons that were learned doing this IoT demo. And then the other one is a related paper.

It's called it's the use case abstract that goes with it and it details all the use cases. So from a proof of concept standpoint, I think sort of mission accomplished, right? We know it works. Right. Now, I'm no longer with IBM, so I can't tell you what the plans are, but I can tell you being here in San Francisco and in Silicon Valley, the number of companies that are taking core pieces of Bitcoin and blockchain technology and figuring out how to scale them and adapt them into billions of devices is enormous. I saw a statistic the other day saying that an estimated $600 million is being invested in. in blockchain-related startups here in the US. Yeah, yeah, I know this is, and I guess also 21 Inc. know, the company has raised like 112 million or something like that.

I guess they haven't really said what they will do, but it definitely seems to involve some sort of hardware component. And so, and I remember there was some mention as well of the idea that, Well, first of all, hardware, mining, and mass market, right? So if you take these three together, that sort of makes no sense in the current mining environment because it just makes no sense to create a hardware device that mines Bitcoin for the masses because you have a complete move towards centralization because the economics are just that way.

But in the sort of context of Internet of Things and what we've been talking about today, maybe one can sort of understand how these three things could go together. I don't want to speculate on what 21 is doing because I really don't know, but I really am, I'm very optimistic. In talking with lots and lots of Bitcoin and blockchain entrepreneurs, I would say increasingly the shift is on how to use this tech. The focus is on how to use this technology to enable other processes rather than sort of the purely focused on just a transactional mining in a, in a digital. currency. Is there any source code out there for ADAPT or anywhere people can contribute or even use this for their own project? I'm not sure what IBM has chosen to release. And unfortunately, I can't give you, as I said, I'm really just here representing my own views. I'm not with IBM anymore.

So I don't know what their specific plans are. I think they've done some really cool stuff so far. I was really proud to be a part of that. And in the work that I'm doing here in Silicon Valley, with some of the companies that I've been talking to, the demand for this kind of capability is enormous. And we see that in startups and also in the take-up of platforms like Ethereum. Do you know of any other initiatives by any other companies who are trying to tackle the same sort of problem and with similar solutions? I do, but I've signed a lot of NDAs.

Okay, well, I guess I'll have to have you in a little bit to talk about them once they're... when it's released to the public. So a listener had some interesting comments in the Lusop Bitcoin Forum because I asked if somebody has some questions on this.

And one of the questions was like, so if you're a developer interest in this area, what kind of, what should you focus on? Like what kind of things are to keep in mind, I guess, so you don't build something that will, you know, not be future-proof? I think this,

single biggest opportunity for investment in the future is around development tools. Bitcoin is a new platform. At IBM, we sort of self-taught ourselves everything. Same thing with the Samsung engineers that we work with who did an amazing job. But it's hard. And as much, if we think about how existing infrastructure is built today, we're all building the same stuff based on the same set of open source components and the same set of development tools, it's much, much harder to build the same capability using the blockchain or Ethereum. So we need a lot more development tools right now. I would say that's the number one opportunity in the short run is development tools. Yeah, absolutely. No, I mean, I actually work

in the Ethereum office here in Berlin and sort of I've witnessed that to some extent. And I know some people are trying to build actually complex applications on top of the theorem now. And it's a hugely difficult thing apparently. It's not easy. I think it will be incredibly rewarding as we get there. But right now you can still tell how much of a frontier business this is. Absolutely. So how far do you think we'll get in the next three years?

That's a great question. I think two or three years from now we will have development tools. It will be typical for people to start building distributed applications and we'll be talking about, we'll be talking about kind of platform competition between traditional cloud services and distributed cloud services. Okay, cool. Do you think, well, this is something I was thinking about is how do we make sure that connected devices will be talking the same language.

So let's imagine that ADAPFORS isn't to this really sort of nice toolkit for on which you can build, I wish companies will be building their infrastructure. And then maybe perhaps, I don't know, Intel or some other company comes up with a different protocol that uses similar technologies but isn't quite compatible. Will we get to some, do you think we'll get to some point where the there are multiple technology stacks sort of competing against each other, and we're running to similar problems where there's incompatibilities between them.

Definitely. We're going to have that. The marketplace is going to ultimately decide which ones are the most attractive. We have that today, right? We have so many different variations of Unix and Linux. We'll have the same with the blockchain. There'll be multiple blockchains and development environments. In general, I would say that the market has,

it seems like the technology industry has shifted from a time when standards were set and then technology was deployed. Now what tends to happen is standards get set quite a few years after the marketplace seems to have voted on whichever ones they like the most. Yeah, but I mean, so I guess that means it will be, will once again be in this situation where you have to choose your ecosystem or people will have to develop other technologies on top of that so that all these standards can talk to each other. Yeah, although I, I, I, I don't think that'll be all that bad of a situation. This industry is incredibly adaptive.

And even though there's lots of different versions of things like Linux, ultimately it's pretty easy to move between them. They're 90% the same. So I'm pretty optimistic that at least in the short run, all this variability will produce a lot of innovation rather than more confusion. Cool. Well, Paul, thanks so much for coming on today.

Is there something else you want to mention or? I've got, I'm going to be getting back into business very shortly. And so watch my Twitter feed at P. Brody for an announcement in the next couple weeks. But other than that, I'm just delighted to have been invited here. And I appreciate your time. And it was a great honor. You know, absolutely. I mean, thanks so much for coming on.

That was really, really fascinating to talk a few at that. And I think it's something that, you know, we sort of were aware. I guess of a certain intersection between internet of things and the blockchain, but certainly not at all to the extent, at least personally that I feel right now, it's like, oh, this is actually could be an amazing, amazing fit and amazing opportunities. So my impression very much is that I guess this is a topic that will come to revisit often, perhaps also when there are more mature actually. So people perhaps develop being blockchain cryptocurrency protocols specifically for that.

I mean, I guess maybe Ethereum to an extent is that, but I'm sure they won't be the only ones. I really share your view, Brian. I think that any time that we get to talk about really practical things, I really feel like this is sort of the somewhat coming to fruition of all these things we've been talking about in the last year, and especially what Mike Hearn has been talking about with autonomous cars, etc. were with this ad-up technology starting to see how that could actually be applied to real life. And that's cool. So yeah, so Paul, thanks so much for coming on and all the best with your new ventures and endeavors.

and yeah we'll look forward to perhaps having you back on at some point to get an update on what's been going on in this area terrific thanks very much and thanks for listening we'll be back next week with another episode if you want to you can follow us on Twitter with episode and yeah thanks and we'll be back soon