In this episode, the host becomes the guest. Sara Drakeley interviews Henry Holtzman, Chief Innovation Officer at MobileCoin. Sara probes Henry's career at the MIT Media Lab during the halcyon days of computing, and Henry shares insider tips on how to design successful tech products and bring them to market. Sara and Henry also dive into stablecoins, which Henry divulges is MobileCoin's next big venture. Get ready for MobileUSD! They discuss the challenges of building out this technology, the difference between algorithmic and asset-backed stablecoins, and what can go wrong with each. Henry gives an expert breakdown of the TerraUSD collapse and what it could mean for the future of stablecoins.
Speaker 1 (00:06)
Hello and welcome back to Privacy is the New Celebrity, a podcast about the intersection of tech and privacy. I'm Sara Drakeley, Chief Technology Officer at MobileCoin. And today on the show, we have one of our very own hosts in the guest chair. Henry, thanks for agreeing to be here with us.
Speaker 2 (00:21)
Hi, Sara. Thank you so much for having me.
Speaker 1 (00:23)
That's Henry Holtzman, the Chief Innovation Officer here at Mobile Coin. And the reason we're flipping things around today is that Henry has a long and storied history in the world of technology. He spent time at MIT with the Media Lab. He was vice President of the AI center at Samsung, and he has founded and launched his own companies. And now we're lucky enough to have him here with us. Henry, how does it feel to be in the interview seat?
I have to admit I'm more comfortable on the other side of the microphone, but it's always a pleasure.
Speaker 1 (00:54)
So let's start at the beginning of your story. I'm curious about the first time you became interested in technology. Do you remember what it was that made you gravitate towards tech?
Speaker 2 (01:03)
Oh, wow. I mean, thinking back, it was around the fifth grade, so it was quite some time ago, and I can remember some of the things I did. I'm not sure I remember what was the shiny butterfly moment that I wanted to chase after, but I got really interested in things, like I made my own calculator out of parts I bought at Radio Shack. And I remember figuring out how to wire up my fifth grade classroom so that everybody had a phone and we could call each other.
Speaker 1 (01:34)
What kind of phone?
Speaker 2 (01:35)
They were like old Rotary dial phones. And that was a lot of fun. And then somewhere 5th, 6th grade, I found out that there was a "learn to program computers" class being offered by the high school as a Saturday thing. And I lobbied really hard to be allowed to take it, even though I was only in the fifth grade. Or maybe it was the 6th grade at that point. And I did not manage to do it. But I did get a hold of all the manuals for the computer. I read them all, and then I would take turns pretending I was the computer and programming the computer. So I spent those two years until I was actually allowed to touch the computer practicing for that day.
Speaker 1 (02:21)
When you are imagining that you're the computer, can you tell me more about your mental model of that?
Speaker 2 (02:26)
Oh, like, I would like maybe have a pad of paper and I would like be keeping track of what was in each memory address. And then I would scratch it out when the instruction was to replace it with something else. And just like I'd read the architecture of how the computer actually worked, so I would just simulate it on a piece of paper.
Speaker 1 (02:45)
And what was it like being at MIT in the 80s? That was such a critical time of growth for computing. And you were right there in the middle of it.
Speaker 2 (02:53)
Yeah. MIT is like an amazing place in terms of, like, the environment, the people, the facilities, the electricity of ideas. I got really excited about my classes at first. I'm hesitant to brag, but I had a straight A after my first year.
Speaker 1 (03:15)
That's really hard to do. That's really hard to do.
Speaker 2 (03:19)
And I took ten classes and I got an A in every one of them in that first year, my GPA from then on, like, slowly went down. Went down, went down to the point where if it was a five year program, I might not have been allowed to graduate because I got interested in everything but my classes. I kind of, like, decided the classes were a bit too easy and started spending all my time doing things like building a radio station, doing computer programming, stuff like that, playing sports, just everything but the going to classes.
Speaker 1 (03:58)
I've even heard a rumor that you opened for Nirvana once.
Speaker 2 (04:02)
Well, I'm not going to deny it, but that's kind of, like, right place, right time. We booked Nirvana, and when we booked them, they could still be afforded by a student group booking a dorm concert.
And then by the time they actually came to play, they had blown up. Like, we went to see them in clubs the two nights before to packed sold out shows. And to this day, I'm amazed that they didn't somehow figure out a way to back out of that contract. They came and played at MIT for, like, hundreds of dollars.
And then, yeah, I got to open for them.
Speaker 2 (04:39)
I got to meet them, hang out with them. It was fun.
Speaker 1 (04:42)
Well, and it's also, I think another rock star status group that you were with was the Media Lab. And I think that for people who aren't familiar with the MIT Media Lab, can you describe what they're up to?
Speaker 2 (04:55)
Yeah. So the Media Lab was an idea that began in around 1979 ish and it was a recognition that at a prestigious university like MIT, professors got their start and got their tenure and got their careers by advancing their discipline. And that meant that they needed to head towards the central problems of their discipline. So if you're like a physicist, you got to do hardcore physics. If you're a mathematician, you got to prove yourself hardcore math. But in the areas of innovation that were really starting to drive society and business, it was cross pollination is where the really interesting things that were happening. And so there was no home for a bright new academic who wanted to work on those kinds of innovative ideas that cross the disciplines. And so Jerry Wiesner, who was the President of MIT, formerly Kennedy science adviser, was approached by the first director of the Media Lab, Nicholas Negroponte, what do you want to do when you're done? And his answer was, I don't know. And so they schemed the Media Lab as the thing to do. And a whole academic program that would go with it, media arts and Sciences.
Speaker 2 (06:15)
And it actually ends up being kind of like right there at the beginning of the Internet. Right. Beginning of the Internet cross disciplinary. And that's what the Media Lab is known for, is having all of these ideas about what the Internet should be a decade before it was those things.
Speaker 1 (06:32)
Yeah. The Media Lab is a really incredible place and something that I also was drawn to. And there's a crossover of human computer interaction as well. There there's the camera culture group who's working on trying to solve light bending around corners and all these really interesting ideas. And I think that then having had that experience there in that place, with all that innovation, you ended up joining Samsung, where you've designed many incredible products over your career, including the Frame TV. I'd love to get your take on what goes into good design. How do you approach designing tech products?
Speaker 2 (07:12)
Yeah, absolutely. I can tell some of the story behind that one. I am going to say that we had some people that I worked with who are actually design is what they do for a living, and I would give them the credit for the design. I was an executive, which means that people had to take my notes, but I didn't necessarily do the work. I don't want to take credit for other people's work there. But yeah. So the process we use there was to start out by talking to people. The mission we got from the company was there's a clear recognition that televisions were, like, defined by the past. And the company wanted to think about what would be the displays in the home that were organized around the future. Okay. And so that was what this lab I created for Samsung, that was its mission. It was like the next display Experience Lab. And so we started out by doing ethnographics, by talking to people, by looking at how they live their lives, by looking for unmet needs in their homes that we could address. The frame came about because we discovered that kind of there were two classes of people when we went in and we talked to them about their TVs.
Speaker 2 (08:32)
They were the kind of people who had to have the latest, greatest TV and it defined their personality. And like, they were the ones who brought all their friends over for football games and stuff like that. Okay. And then there was this other set of people that were embarrassed about the TV. It was an eyesore. They didn't want people thinking they watch TV all the time, and they were looking for pieces of furniture that could hide their TV. They wanted to talk about, can I use a projector that I can roll up and make disappear. And so there was this realization from that research that Samsung kind of had nailed the first set of people. We had these big, beautiful, curved TVs that everybody wanted to show off. But for that second set of people, we didn't have anything. And so we embarked on this question of how we transformed the TV from an appliance to decor. And we came up with a bunch of ideas for how that might happen. And one of them was this make it look like art on the wall when you're not watching TV. And that's what the frame is all about.
Speaker 2 (09:35)
It can very convincingly, through a few really simple but clever technology tricks and some masterful industrial design. It looks like a framed picture on your wall. And people are genuinely fooled when it actually turns out it's a TV and will play a TV show.
Speaker 1 (09:54)
I'm going to draw a comparison here. Bear with me, because this may not fit exactly. In cryptocurrency, I think there's similarly a divide of people where there are some people who are really excited about the fact that they're using cryptocurrency for different products, they're excited to tweet about it, to tell their friends about it, to show them the new product built on top of cryptocurrency. And then there are other people who feel less enthusiastic about it, or even confused or even angry. Why? This is a new technology approach that may not have products that are quite solving the use cases that people need solved. And I think that was something that your team did at Samsung is identify that there was this use case for a TV that was art and that solved this actually really important problem that people had with TVs. Are you seeing similar things like that in the cryptocurrency industry, where there's a bifurcation of users and how they're thinking about cryptocurrency products?
Speaker 2 (10:55)
Yeah. I mean, there's definitely a lot of different camps in the cryptocurrency world. I would say, if I look at it from the lens of MobileCoin so far, cryptocurrency is something that people are thinking of as an investment primarily. Right. How much will my coin be worth tomorrow, the next day, next year? Also, there was kind of this utility idea that came about with, like, Ethereum. Ethereum is actually the thing that hooked me into the modern crypto wave, because I saw the Bitcoin thing happening and I kind of kept thinking to myself, yeah, but what's it actually good for? Then Etherium came along and there was this notion of being able to modernize how contracts work, being able to encode rules and future payments and things like that, like if you're an artist. Right, and you want to sell your piece of art. And we've all heard of, like, NFTs. One of the great things about NFTs, which you can think of as like a certificate of authenticity. And one of the great things about the NFT is that the artists can choose one where whenever the owner of the art resells it, they get a percentage of the proceeds.
Speaker 2 (12:09)
And that's what Smart Contracts bring to the world is being able to plan for the future in a way the code will create that future. That's pretty cool. I got excited about that. And the first crypto I ever bought was Ethereum. For us at MobileCoin, the thing that I think we're really excited about is actually helping people with digital cash. Like, how do you like, that's our utility, our utility is payments, and we want payments that are as easy to do as cash, maybe easier. Like, I can't give you cash if you are 6000 miles away, right. But I can send you some mobile coin if you're 6000 miles away. We want it to be private, though, like cash. We want it to be quick. All of these things are not what you get from a Bitcoin or Ethereum. Now, there are other crypto projects kind of wanting to do that as well. So we're not unique, but this is like a use of crypto that is fledgling. It's not the thing that most people out there who are following the crypto news are doing with crypto.
Speaker 1 (13:22)
Yeah, I have found that as well, where the industry almost seemed to give up on crypto as being a viable payments technology. And I think that's something where MobileCoin stayed heads down on, let's solve this problem specifically for messaging apps. And now we're popping our heads back up. The rest of the industry went to DeFi and NFTs, and I think there's a real place for this use case, and it's exciting.
Speaker 2 (13:52)
We can pull this, like, full circle to the MediaLab discussion because I can't even remember what year. But one of the earliest cryptocurrencies that people started to hear about was something called DigiCash. And that first boss at the Media Lab, director of the Media Lab that I was talking about earlier, Nicholas Nicco Ponte, ended up being the CEO of DigiCash.
Speaker 1 (14:14)
Speaker 2 (14:15)
And we got really excited about it. We got excited about this idea that our money was going to go digital, because that's actually what the Media Lab was about. It was about turning atoms into bits of all kinds. And the notion of that cross pollination I was talking about, it really hits its stride when everything's on a digital medium, suddenly the physical world is no longer a barrier to mixing together businesses, disciplines, industries. Right. And that's what we were all about. And to get that to happen for money was really exciting. And plus the promise of doing it in a cash-like way, not just like numbers on ledgers. Unfortunately, it was too early. It didn't go anywhere then.
Speaker 1 (15:02)
Another aspect of the DeFi world is stablecoins. And I think that it's been in the news recently, and we could go at length into stablecoins. But I'd love if you could take a quick moment to describe what is a stablecoin and what's its place in the cryptocurrency ecosystem.
Speaker 2 (15:26)
Right. So as I was talking about before, we have these assets that people are buying, like Bitcoin, where they're buying it because they think it will be worth more in the future, but sometimes it's worth less. Right. So if I buy some piece of crypto in order to pay you, and probably we've decided not that I owe you a certain number of Bitcoin, but that I owe you like $100. Right. And if I buy $100 worth of Bitcoin and then I send it to you and then you're going to sell it and it's only worth $90, you're like, hey, Henry, you still owe me $10. Right. And so this volatility is going up and going down of these crypto assets that gets in the way of using them for payment and enter stablecoins. And just that word stable, just the whole idea that they're going to pretty close to track the same value all the time. Now, there's a couple of ways that crypto projects do this. One is they put actual dollars in the bank. Okay. So if I'm going to issue you a Henry dollar, I'm going to take a dollar for you and put it in my bank account.
Speaker 2 (16:38)
And I'm going to make the promise that anytime you're tired of having that Henry dollar, you can give it back to me and I'll give you back your dollar. And that's one way people do stablecoins. Some of the well known ones that do that are things like Tether and USDC. There's also ones that are called algorithmic stablecoins, and those have some kind of complicated system for keeping collateral and selling the collateral in order to stabilize the stablecoin. So they're basically being done by algorithms. And we just recently saw one of those collapse, and that was Terra USD. And there have been a lot of high profile collapses of algorithmic stablecoins. And there are also some that are still doing okay.
Speaker 1 (17:26)
In the cases where the algorithmic stablecoins experience collapse, can you describe what happens, maybe even specifically for Terra?
Speaker 2 (17:34)
Well, I think we're still close enough to the Terra collapse that we're not exactly sure yet what happened or why what happened happened. But one of the theories is that to get people excited about Terra USD, they were offering, if you took your Terra USD and put it in a special kind of account, then you would get interest. In fact, a lot of interest, 20%. Okay. And something like 80% of all the Terra USD was in one of these accounts or another. So that's what people were doing was they were buying up Terra USD and they were depositing it into these accounts in order to make the 20% interest. This drove a demand for Terra USD. The way Terra USD worked is that when you wanted to make more of it, you would sell a coin called Luna. And selling this coin called Luna would cause some of the Luna to get burned, some of it to be put into an account, like sequestered. And so in general, there would be less Luna in the world. And so the price of Luna would go up. And so there was like this very nice cyclical action of Luna becoming worth more.
Speaker 2 (18:49)
They're becoming more Terra USD, more people earning money on the interest of staking. They call it staking when you put it in those accounts, staking their Terra USD. And then I think there was like a rumor that they weren't going to be able to pay the 20% anymore. And that was coupled with people who are already scratching their heads and saying, where is the money coming from? That's paying this interest? And I think maybe some news started to surface that sort of like investor money was being spent in order to provide that; there wasn't like a sustainable source of this interest. And so people started to sell off their Terra USD. They took it out of the staking, they started to sell it. And as they did that, the price of the Terra USD went down because the system couldn't instantly absorb all of this terror. It's kind of like a supply and demand thing, right. It takes the algorithms a little bit of time to react to all of that. And so as the algorithms tried to react, the price of Luna started to crash, going the other direction. The more Luna crashed, the more Terra USD crashed until it got to a point where it's hard to imagine that it can ever recover.
Speaker 2 (20:00)
I don't know what Terra USD is selling for right now. Luna was a very small fraction of a cent last time I looked. And I think Terra USD may be in similar shape.
Speaker 1 (20:13)
Yeah. It's really unfortunate when those events happen. And part of it is being in this innovative space where we're trying to innovate across so many different directions. And part of it is risk management and really trying to take a larger view of the ecosystem, the way that supply and demand may affect things where you might end up in a complex spiral of consequences.
Speaker 2 (20:42)
So I think one of the interesting things about the dollar-backed stablecoins is that there's generally two ways to buy or sell those. One way is there's just crypto exchanges where you can go and buy it or sell it. Okay. And those crypto exchanges, the price actually floats, right. So if there's an excess of supply, it goes below a dollar. If there's a shortage of supply and an excess demand, the price will go above a dollar, just like at any marketplace. Right. But there's a second path. The second path is you work directly with the company like I described before when I said, if you ever get sick of your Henry dollar, you give it back to me and I give you a dollar. So usually those companies have a route where you can actually deal with them directly and get your $1 for your $1 of coin. And so during this whole Terra USD crash, there was a moment where Tether started to have this problem, where on the exchanges it went down, like in the 90s into the low $0.90 as people started to panic and want to sell that one as well.
Speaker 2 (21:51)
But the CTO there sent out a tweet reminding people, if you want a dollar for your tether, you just have to come to this website and we'll get you your dollar. It's a slower process. It involves bank transfers, stuff like that. But you reminded people that was available. And then over the next hours, day, whatever. I read later, $7 billion of Terra, that's a big number, 7 billion were bought up off of the exchanges, reducing the supply, bringing the amount back up and redeem directly to through Tether. And they were actually able to keep up with that. Like, they fulfilled that promise of giving back the money to the tune of $7 billion over the course of a fraction of a day.
Speaker 1 (22:41)
That's incredible. And we are also thinking about developing a stablecoin at MobileCoin, and we're learning from the industry from what's working, what's not. Can you describe a little bit about what a MobileCoin, stablecoin might look like?
Speaker 2 (22:56)
Yeah. First of all, I just want to again, I like to connect the dots. When we first brought MobileCoin onto the market, we got a lot of feedback that people were concerned about using a coin for payments that was volatile. They were worried about the price of MobileCoin being different every time they looked. So that was some early feedback we got about our product and using it for payments. And so since that moment, we've been like going, okay, let's deal with that. And we've been working very hard under the covers to deal with that. The first step in that for us is to take our blockchain, which is what they call a layer one blockchain, which means that we have our own native protocol for how our crypto works. So we're not like built on top of Bitcoin or built on top of Ethereum. We are our own thing. It was designed for having a single token, MobileCoin, which we call MOB. And so the first step was to add the ability to have other tokens, and we're really close. And so we get to a stablecoin by being able to introduce a second token.
Speaker 2 (24:13)
We actually made enough in making this change. We made the ability to make enough tokens that I think are, I don't know, how many items are there in the universe? We may be able to make a coin for every atom.
Speaker 1 (24:29)
Unfortunately, atoms are not stable.
Speaker 2 (24:31)
Yes, that's true. But we're focused on our next one, which we're calling Mobile Dollars being a stablecoin.
Speaker 1 (24:40)
Great. And when you're thinking about all of the new tokens that could be on the blockchain, as we're evolving this protocol, does that degrade the privacy anyway, or how has MobileCoin approached that, since privacy is so important to us?
Speaker 2 (24:54)
Yeah, great question. Great question. So anybody who's not familiar with MobileCoin, probably all you listeners are, so I'll be really brief about this. One of the unique things about our blockchain and why we made a blockchain from scratch is for all of the transactions, when you look at them on the blockchain, you can't tell who sent the transaction, you can't tell who got the transaction, and you can't tell how much the transaction was for. That's all between the parties that were involved and not for prying eyes. But the blockchain does keep everybody honest. Okay. So in putting this new field into our transactions, the token ID, we decided we wanted to do the same thing. So in this yet to be released new version of the MobileCoin blockchain, now there will also be a token ID. That's the kind of token being transacted, but that will also be confidential.
Speaker 1 (25:45)
Great. So I could send a transaction with potentially, because transactions have potentially multiple inputs, multiple outputs, and I could even mix different token IDs in one transaction.
Speaker 2 (25:56)
So in our upcoming release, we're not quite there yet, but that's just in the engineering pipeline right behind.
Speaker 1 (26:05)
Fantastic. And in MobileCoin transactions, we have this privacy of the sender, receiver, and the amount, as you've described. And is that not possible on Bitcoin, what's the difference there? What can you tell on Bitcoin?
Speaker 2 (26:18)
Yeah, on Bitcoin, everything is transparent. Okay. And often I think when people hear it's a cryptocurrency, they think about the encryption, but the encryption isn't hiding anything in Bitcoin. What the encryption is doing is. It's like signatures, it's keeping people honest, but everything is transparent. And so, for example, if you send me a payment, you have my address and you can now actually go and look on the blockchain on the Bitcoin blockchain and figure out how much I'm worth, who else I do business with, and maybe even from some of those things where I've been.
Speaker 1 (26:59)
Actually, I think this is a confusing point for a lot of people, because the beginning of Bitcoin, there was this idea that it was anonymous and it was pseudonymous. And I think that what we've seen prove out is that a lot of the assumptions around how that anonymity worked proved not to be true. And as more interactions happened. And if I paid you, Henry, then I have your address for all time and I can see what's going in and out of there. And I think that wasn't necessarily obvious to start. And now that we have all this data, we've seen it being used for years. We see the ways in which it's not actually anonymous and that there needs to be some evolution there. So while you've been at MobileCoin, you've been working on lots of different projects, including the stablecoin now. And what have been the challenges that have really gotten you out of bed in the morning?
Speaker 2 (27:46)
That's a great question for me. It's an everyday it's a different kind of thing, which is part of what's exciting about it. I really enjoy doing new things all the time. So I really do get out of the bed every day ready to solve yesterday's problem. If I can think about what some of those big problems have been, you know, they don't actually, they're not things I dwell on. It's like, you know, like, solve the problem, move on, solve the problem, move on, solve the problem, move on. I think a big problem that we continue to work on is adoption. Right. It's kind of like there's the classic thing. People talk about this a lot. Like, let's say you invent the Fax machine, who's going to buy the first Fax machine? What can they do with it? But once everybody has a Fax machine, you don't want to be the one person without one. Okay. And so creating a new currency is kind of like that. And I think that is a thing that does continuously. I think it gets us all up in the morning because I think we're convinced that we have something of great value to people.
Speaker 1 (28:56)
When we think about adoption. That is a perpetual problem for not just all the cryptocurrency projects, but also in lots of tech companies that have products where they're trying to get more adoption for these innovations and tools that they think are going to help users. And how is MobileCoin thinking about that problem? Or how have you thought about that problem? Also when we look back at the Frame or other projects that you've worked on?
Speaker 2 (29:21)
Yeah. So I think I usually start from a place of trying to provide value to the user. Like when I think about innovation. Right. Like, the Holy grail of innovation is value and viability coupled with feasibility and novelty. Those are the four things that make something an innovation. It's new, it can be made, people will find it valuable. And the company that's doing it has some sort of incentive, like it's viable, sustainable. Okay. So I like to start with the value. What are we doing for the user? Find something the user will find valuable. And I like to also think about as we achieve scale, how do we stay, like, with the user at the heart of what we're doing? We can look like in the environment now. We can see all kinds of companies that have gotten past the scaling problem, so much so that, you know, now the government may be going after them for having monopolistic control of certain things. Right. So, like, you know, you take a look at, like, social networking. What if I wanted to create a new social network, I have no chance. Right, because that space is locked up by a company that has everybody already.
Speaker 2 (30:52)
I think as we look at those companies, we ask ourselves, we see the tremendous value we get from these companies. Right? But we ask ourselves now, who are they doing it for? Are they still doing it for the user or are they doing it for the company? And I think we have our doubts about their altruism. I think we have our doubts that they're making their decisions from day to day about what's best for the user. And instead they're making that decision day by day as what's best for the shareholder. And they might try to turn that around in PR and say, well, of course, what's best for the user is best for the shareholder because we only have value if users continue to use us. But then we have whistleblowers who reveal memos that show that that's not true at all, that the algorithms are optimized for the revenue, not for the health of the user. So that's my sincere desire for the future of MobileCoin is we find that scale that is viable, sustainable. But we hold on to the idea that what we're doing always has to be in service of the user's best interests.
Speaker 1 (32:08)
Am I correct in hearing that you are an optimist for how technology can serve the user?
Speaker 2 (32:13)
Oh, yeah, I am very much a techno optimist. I mean, I'm also like a gadget head. I see all of the perils and then I think about them and decide not to completely turn away from these technologies before they've proven they're ill is premature. So I'm very excited about what technology has done for us up to date and what it will continue to do for us. I really believe in it. But yeah, we have to have our eyes wide open because there are perils and we're seeing a lot of them.
Speaker 1 (32:51)
And now get ready because I'm about to ask a question that you should be very familiar with. It's a question you like to ask a lot of our guests, but I'm excited to hear your answer. When did you first realize that privacy was important to you?
Speaker 2 (33:04)
You know, I think I probably, like every child, had some thoughts about this in grade school where I felt like my parents were trying to control me and stopped me from being me and wanting to make me into something else. And that the way to cope with that was just to hide, hide from them, to find some privacy. I have thought back to those times frequently, and I think I was pretty aware, even back to being a kid, how important privacy is. Professionally, there have been many moments along my professional career where I've realized that privacy was out of whack from technology and that it was my responsibility having been somebody who kind of like squirted that toothpaste out of the tube to do what I could now to focus on how to bring more balance into the privacy situation. And one of those was Internet of Things. I did a lot of early work on Internet of Things, and I even started a company that was one of the very earliest IoT companies. And we went out to the Fortune 100 companies, and we built a demonstration for them where if they added just a little bit of radio technology, well under a dollar, 10 cents, to their product, it now became a little bit Internet aware, and the Internet became a little bit aware of their product.
Speaker 2 (34:24)
And now we could knit whole new experiences for their users that actually were focused around their product instead of around finding a computer. We're talking, like, for context, 1998 ish here. Okay. And so, for example, we did a thing for Kodak. It's just a demo. Like, they weren't paying us to do it. We were trying to get their business, but we did a thing for them where we embedded one of these little circuits into every print that they made. And then if you took that print and waved it over your mousepad, it would pull up the photo album that that print had come from, giving you the opportunity to see all the pictures and context. So those were the kinds of experiences we were building. We did another one with a hotel chain where they were known for you feeling like you'd come home. And part of that was being recognized. And their highest paid employees were actually their doormen because once they hired them, they wanted them to be there for decades and to actually recognize the regulars and say hello to them. So, again, this little bit of radio technology, we put it in their luggage tags, and they had a bracelet that was wired up to a little earphone.
Speaker 2 (35:37)
And when a customer came back, it would whisper into the valet the name of the customer. And so they could greet them by name. And their customers loved this kind of behavior. But as we went to these Fortune 100 companies and showed them our ideas for their technology, they came back almost to a one wanting to spy on their customers, to get them to buy more of their services or to get more money out of them. Like, they kept throwing at us ideas that just immediately I had a reaction to it. You want to spy on your customer kind of thing. And so when I got back to MIT, I was doing this on a leave. The next phase of my work at MIT was making Internet of Things into being a privacy preserving Internet of Things.
Speaker 1 (36:23)
My mind is a little bit blown about a privacy preserving Internet of Things. Yeah. And how necessary that is, because I know I'm sure you heard lots of the stories around being able to read Internet traffic from your neighbor because they're using a hue light or something, you end up exposing a lot more surface area.
Speaker 2 (36:43)
Yeah. And so you'll enjoy as a cryptographer, one of the things we did back then, this is like maybe 2002 ish, is we added to these little radio circuits. They had an Identifier that they would give out. We added a hashing function and a seed. Right. And we would rehash with every blast. Okay. And just keep hashing the hash. So we have this hash chain. And so the thing actually never gave out the same identity twice. And in order to be able to correlate those hashes, you needed to be somebody who knew the seed.
Speaker 1 (37:24)
Like a little RFID ratchet. Yeah.
Speaker 2 (37:26)
Yeah. That's exactly what it was.
Speaker 1 (37:28)
So I think for people who are not familiar with technology, they see a bunch of data that's like bytes, and it's really hard to Intuit what that is if you're reading the data going by, but the computer knows what that is. And I think there's also confusion. And maybe this goes back to the misunderstanding of privacy at a computational or statistical analysis level, the difference between compression and encryption. And I bring this up because I know that you know a lot about compression that I would like to touch on a little bit. And how do you think about the ways in which we can represent data and some of the maybe non obvious results of how we represent that data?
Speaker 2 (38:12)
Yeah. So in compression, we do talk about two styles of compression. We talk about lossy and lossless. And the idea of lossless compression, also called redundancy reduction, is that you figure out a way to represent the data so that it's just in a more compact form. But you can get back the original data anytime you want. For example, let's say your data tends to have few ones and lots of zeros. Then you might do something. It's called run length coding, where you instead of writing zero, zero, zero, you say ten zeros in a row followed by a one, followed by 15 zeros in a row, followed by a one. So by talking about how many zeros in a row, you can represent it much more compactly, but still be able to write out the original pattern of one zeros. The other kind, the lossy kind, is also called irrelevancy reduction. You can never get back the original, but you don't care, because in going from the original to the compressed form, you preserved everything that anybody cared about. Okay. And this is what we do with images, like a JPEG is we do actually some redundancy reduction as well.
Speaker 2 (39:29)
But a lot of the magic there that gets the big wins is we figured out that the human eye just isn't that sensitive to certain things. And so we make a picture that looks like the original picture but actually has much less data in it.
Speaker 1 (39:43)
And you were part of setting the standard for MPEG4. Is that correct?
Speaker 2 (39:47)
That's where we were going. Yeah.
Speaker 1 (39:49)
Speaker 2 (39:53)
I am so into compression because of having worked on image compression, in particular, video. So MPEG, which is used on everything from YouTube to TV, broadcast TV, DVDs, Blu Ray, it's like basically how video is transmitted. I got to work on that right from the start of the standards.
Speaker 1 (40:18)
So it's easy to think about the privacy abuses enabled by tech like surveillance capitalism and data tracking and all of that. But solutions are always harder. What do you think the role of technology should be in society?
Speaker 2 (40:32)
Frequently in my career, when I've thought about these conflicting forces of getting the good out of technology and the privacy incursions, I have not thought that the answer was to regulate away the technology, but rather I think there's a large role for regulation for laws and that those laws should focus on the privacy incursions themselves instead of on the technology that might be causing the privacy incursion. So what I would like to see is for legislatures to really strongly create a regime of Privacy, a right to Privacy, and then no matter what the technology is, they should describe when a company, an entity, is violating that right to Privacy, what that looks like and what your remedies are, because technology tends to be a little bit neutral. Right. In terms of is it for good, is it for evil? It depends on how it's used. So let's not try to regulate away the technology. Let's regulate the use. That's my belief in how we should approach that kind of stuff.
Speaker 1 (41:46)
All right, Henry, time for the most classic of classic questions. Privacy is the New Celebrity. What do you think?
Speaker 2 (41:54)
Yeah, we ask that of all guests, I think. So it's fair for it to be asked of me. I think I recall that meme going out maybe when Sean Parker started saying it. And I think Sean was talking about scarcity that once upon a time, celebrity was scarce and celebrity could equal economic security for those that knew how to use it. And so people desired celebrity, but it was really hard to get. And now we live in a world through social networking, through Internet media, where celebrity is actually something that is more Warholian. Everybody has their 15 minutes of Fame, but Privacy is more and more scarce. So I see that statement through that model. I think that's a really keen observation is we've gone from celebrity being a scarcity to Privacy being a scarcity.
Speaker 1 (42:55)
Henry, I think we're going to leave it there. Thanks for being our esteemed guest on another episode of Privacy is the New Celebrity.
Speaker 2 (43:01)
Thank you, Sara. It's been a real pleasure to be here.
Speaker 1 (43:04)
How did it feel being on the other side?
Speaker 2 (43:06)
Well, I always like to talk about myself. So good.
Speaker 1 (43:10)
We're happy to have you on. Thanks for listening. Don't forget to subscribe wherever you listen to podcasts and check out mobilecoinradio.com for the full archive of podcast episodes. That's also where you can find our radio show. I'm Sara. Our producer is Sam Anderson and our theme music was composed by David Westbaum. And remember privacy is a choice we deserve.