When Balaji S. Srinivasan sat down for his conversation with Tyler he was the CEO of Earn.com. Today he is the CTO at Coinbase, which acquired his company in the intervening weeks (congrats Balaji!). But while his job title has changed, his passion remains the same: harnessing the power of the blockchain to launch a new generation of entrepreneurs, businesses, and entire markets.
Balaji talks with Tyler about the potential of the blockchain and beyond, including how firewalls may become the new immigration policy tool, why drones are still underrated, the future of news and academia, what the Silicon Valley opener reveals about how America views the tech industry, and more.
Listen to the full conversation
Read the full transcript
Note: this transcript has been edited for clarity.
TYLER COWEN: I’m here today with Balaji Srinivasan. Marc Andreessen has described Balaji as the man who has more good ideas per minute than anyone else in the Bay Area. He is the CEO of Earn.com, where we’re sitting right now, a board partner at Andreessen Horowitz, formerly a general partner. He has cofounded the company Counsyl in addition to many other achievements.
On top of all that, being thought of as simply one of the most generative thinkers, talkers, and speakers in Silicon Valley. Balaji, thank you for doing this.
BALAJI SRINIVASAN: Tyler, nice to see you.
COWEN: My first question, I suppose, is about management. There are amazing achievements from the tech sector, but a lot of them seem to cover software or other services that are very rapidly scalable. To what extent do you think the Bay Area has actually come up with a new and innovative management technique for managing smart people better? Or is it just that you do scalability great?
SRINIVASAN: That’s a really good question. I think the most fundamental concept really boils down to equity, which is to say that in game theory — the simplest game, the two-by-two matrix, whether you’re talking prisoner’s dilemma or something else like that — you want to actually have the upper diagonal where it’s a win-win be obviously better for both parties than any other square on the board, certainly the win-lose or the lose-win case.
What equity does, it takes that from the two-by-two to the two-by-two-by-two-by-two case, the 2^k case, where even if you have k parties, all of them realize that their highest payoff is from the win/win quandrant— that is to say, when the company exits, that’s what makes them the most money — than anything that would be an off-diagonal, political thing to do, which benefits them at the expense of somebody else.
So the most fundamental management technique I think that the Bay Area has pioneered is introduction of equity and company stock options as a given.
COWEN: Say we took Silicon Valley or San Francisco managers and brought them into the healthcare sector. Is there a low-hanging fruit that could be reaped from what you’ve all done here?
SRINIVASAN: The healthcare sector’s a very large area. I’d say that if you separate it out into, let’s say, health IT and electronic medical records and then diagnostics and then, let’s say, prosthetics, then you get to therapeutics and gene therapy and vaccines and things like that. That’s kind of a gradation of risk and also gradation of, let’s say, susceptibility to a software-based attack.
For example, health IT and EMR, that’s directly within our wheelhouse of software. There’s a lot of companies that are working on that already. Diagnostics is something I’m also pretty familiar with. It’s about half physical, and the rest of it’s digital.
Prosthetics, again, is like robotics, and even though there’s invasive aspects of it, you can get pretty far thinking of it as a software and hardware kind of thing and less messy biology. It’s when you get to therapeutics and when you get to gene therapy, when you get to vaccines and drugs that are under CDER or CBER, that’s where it gets difficult.
If you break healthcare up into those segments, some of them are amenable to startups and Silicon Valley techniques today, and some of them not for some time.
COWEN: As you know, we live in a country where really large numbers of parents are afraid or unwilling to vaccinate their children. Or maybe they do so with hesitation. How is it that we explain to them they’re going to get something on the blockchain? It’s sinister, it’s inhabited by Satan, it will eat out the interior of their minds. How does this work? What’s the role of the new Silicon Valley CEO?
SRINIVASAN: Well, that’s mixing a bunch of different things together.
I’d say vaccines are a real edge case for folks who are generally “live and let live” because while you wouldn’t want to interfere with someone’s choice of clothes or their choice of food necessarily, if they don’t believe in vaccines, and their child or themselves has a contagious disease, then it can screw up other people’s livelihoods and lives.
I’m not sure that’s something that a software CEO could attack. It’s kind of an education problem. I do think that folks who are against vaccination, if they wanted to self-quarantine, that would be good. Since they’re not going to do so, then it’s a challenging problem. It’s one that I don’t think software really has a solution to at this point.
On designing journalism from scratch
COWEN: If you were designing journalism from scratch, what would it look like?
SRINIVASAN: That’s an interesting question. It would have much, much more involvement of people on the ground and many, many, many more part-time journalists and people who wouldn’t even call themselves journalists.
For example, the person who’s actually on the scene with a smartphone — everybody now has a 24/7 internet-connected video camera with all kinds of software on it. Korea pioneered something like this called OhmyNews a while back. It’s still possibly going. I haven’t followed it very closely.
A model like that on a global scale, where you’re effectively paying folks in lots of different countries to collect news, is interesting. The news problem, of course, is a much, much bigger thing.
COWEN: There’s still verification, right?
SRINIVASAN: There’s still verification, yes.
COWEN: Especially with phony video being easier and easier.
SRINIVASAN: That’s right. One thing that’s a secondary point but somewhat related is, typically what people do is they will take, let’s say, war reporting and Kim Kardashian–type reporting and put them on opposite ends of the spectrum like, “This is your eat-your-vegetables stuff,” and “This is your infotainment.”
I would actually group them at the same region and put, let’s say, tutorials on the other side. Here’s why. Whether it’s about Afghanistan or Iraq or the Kardashians, that’s a story whose facts you can’t independently verify, versus a tutorial on a new piece of software or how to fly a drone. Tutorials are something you can independently verify since you can fact-check every single line by going through it yourself.
What I think is likely to happen is, you’re going to see more news that falls into that news-you-can-use category where, if people are concerned about the veracity, it’s got an inbuilt fact-checking mechanism because you’re actually executing every single line as you read it.
COWEN: This will involve micropayments and the blockchain?
SRINIVASAN: [laughs] There is one aspect that it might. While I certainly think about the blockchain a lot, the part that’s relevant here is prediction markets. While people have often talked about it — lots of people write op-ed columns — there’s no record of whether they were correct or not in their predictions. People who opine for a living don’t traditionally have a way of collecting, “Hey, I was right about this, wrong about that.” Venture capitalists do. It’s called returns. [laughs]
COWEN: That’s right.
SRINIVASAN: You could potentially bring something like that to a wider audience if you nail the whole prediction markets thing. That’s something people have talked about for some time. It’s newly feasible with blockchain. Maybe someone will take a crack at that.
COWEN: Why aren’t private companies more interested in prediction markets? This has long been a puzzle to me.
SRINIVASAN: Several reasons. First is, private companies have the market itself, namely, the stock market, the commodities market, the bond market, where people are paying . . .
COWEN: But that’s conditional on many different decisions.
COWEN: You could have a conditional market: “If we do X, what do you all think?”
SRINIVASAN: Yes, except the problem is that those real markets have real money, and all the folks playing in those markets are really playing to win.
Whereas a toy market — at least that’s what we were perceived as at the beginning — we don’t have the dollar totals required to be interesting. It might only have academic nerds as part of it, as opposed to folks trying to game the system. It wouldn’t be, let’s say, as reflective of reality and wouldn’t have the predictive value as the real markets do. That may be one reason that private companies have not adopted it.
COWEN: But if people value something — take Bitcoin. When Satoshi started it — if that’s even the right word — it was the weirdest possible thing, and that is well underway and then some, after basically 10 years. The iPhone comes out. There’s plenty of markets in iPhones and things that go with iPhones.
Prediction markets — it shouldn’t seem that jump-starting should be that hard. You get liquidity, people want to trade, companies want to do them — just like changing accounting practices or many other things that happen all the time. Do we not want to know the truth within companies?
SRINIVASAN: As I said, I do think people will take another crack at this with the blockchain because the blockchain has not just made it much easier to start your own currency in your drawing room; it has also made it much more easy to start your own marketplace. So the threshold for starting a new market has dropped dramatically. Maybe we’re going to see some innovation in that in the years to come.
On designing academia from scratch
COWEN: If you were designing academia from scratch, what would that look like?
SRINIVASAN: Ah, that’s interesting. I think a lot about the culture of open source versus peer review. The distinction is, in open source, bugs are expected. Reproducibility is a function of your ability to execute the source code. Everything is fundamentally transparent because it’s all published. And you publish first, and then people review it and give feedback second, as opposed to the academic model.
COWEN: Like in the Clay Shirky model, right?
SRINIVASAN: Exactly. If you think about academia, all of those are inverted. For example, a retraction or a correction in academia is a real blemish on your career. It can be a very big deal.
And many, many papers are not at all transparent in their methods or their data. I’d like to see a more open source–like culture. That can really only happen when the incentives of academia are aligned towards open source. Right now, the grants model is not conducive to that.
COWEN: Would you bundle research and teaching the way we do it now, or you have another proposal?
SRINIVASAN: What’s starting to happen is, a lot of the folks in a previous generation who would have gone on to be, let’s say, professors at Stanford are increasingly saying, “Hey, look. Why should I go and spend all this effort for a few hundred thousand dollars for a grant when I can spend a comparable amount of effort, start a company, exit it, and get a few million dollars, and then fund the equivalent of graduate students via angel investments?”
The division between research and teaching comes from basically the 20th century. What you could instead get is something where you start out as a practitioner. Then you do a bunch of angel investments. Then you spend some time researching. And then you go and put up a MOOC lecture, and you teach what you’ve learned.
It’s more organic, where some people will focus just on one thing, some on other things. Some will do a mix. But it’s going to be more commercially aligned, the entirety of it, as opposed to just being in academia.
On why geography still matters — for now
COWEN: Why is the venture capital model so geographically clustered? So much of it is out here in the Bay Area. It’s spreading to other parts of the country. Around the world, you see Israel, in some ways, as being number two, per capita number one. But that’s a very small country. Why is it so hard to get venture capital off the ground in so many areas?
SRINIVASAN: That’s actually now changed with the advent of ICOs and Ethereum and crypto. Historically, the reason for it was companies would come to Sand Hill Road. One maybe slightly less appreciated aspect is, if you come to Sand Hill Road and you get VC financing, the VC who invests in your company typically takes a board seat. A VC does not want to fly 6,000 miles for every board seat if they’ve got 10 board seats and four board meetings a year per company.
What a VC would like in general, all else being equal, is for you to be within driving distance. Not only does that VC like it, so does the next VC in the B round and the next VC in the C round. That factor is actually one of the big things that constrains people to the Bay Area, is VC driving distance, [laughs] because VCs don’t want to do investments that are an entire world away.
With the advent of Ethereum and ICOs, we have finally begun to decentralize the last piece, which was funding. Now, that regulatory environment needs to be worked out. It’s going to be worked out in different ways in different countries.
But the old era where you had to come to Sand Hill to get your company funded and then go to Wall Street to exit is over. That’s something where it’s going to increasingly decentralize. It already has decentralized worldwide, and that’s going to continue.
COWEN: With or without a board seat, doesn’t funding require a face-to-face relationship? It’s common for VC companies to even want the people they’re funding to move their endeavor to the Bay Area in some way, not only for the board meeting. They want to spend time with those people.
We’re doing this podcast face to face. We could have done it over Skype. There’s something significant about actually having an emotionally vivid connection with someone right there in the room. How much can we get around that as a basic constraint?
SRINIVASAN: Obviously, in-person is always going to be valuable so long as we’re physical beings. But I do think that there’s a number of things that have (a) made remote work better and (b) made the Bay Area more disadvantageous. I’ll list a few of the factors. First, soaring rents in the Bay Area.
COWEN: But that’s precisely because face to face is efficient.
SRINIVASAN: True, but at a certain point, what happens is, price is a signal. The ‘rest of the world’ is literally everywhere outside of the Bay Area. There’s a lot of places that are cheaper, that are safer if you’ve got a family, that have better weather. The Bay Area has got decent weather, but there’s places that are nicer, that have a lower cost of living. Very expensive over here.
You don’t have the same issue with engineers getting poached. That’s a big issue at many companies in Silicon Valley. There’s a list of reasons that are starting to pile up that make the Bay Area less and less attractive from a comparative standpoint.
Then on the pull side, remote work is getting better. Every year, it gets better, whether it’s Slack or whether it’s mobile phones getting better.
COWEN: But face to face gets better, too, because Slack is a complement to face-to-face interaction. It’s making both the remote and the direct better. Is it changing the ratio?
SRINIVASAN: Yeah, at least anecdotally. What I found is, a lot of the companies that I’ve been working with have a Bay Area HQ, where your hubs are in-person and your spokes are international. What I mean by that is, it is true that executive functions, management, things like that where you’re managing people, then face to face is good. Whiteboard sessions are good, et cetera.
If you’re really just primarily working with electrons, you don’t need face to face as much, and you can be on your own schedule. Let’s say, if you’re a manager, if you’re an executive, yes, face to face is still probably necessary. But as an engineer, it’s less and less necessary, I think.
COWEN: If you try to think it through, what’s the main failing at the margin of telepresence? If I said, “Well, I think the main failing is you cannot send smells over telepresence,” that doesn’t sound plausible. It seems you could get on with people and work with them face to face, even if you didn’t have a sense of smell.
What exactly is the thing that’s missing if we get into a very good teleconferencing room, we all sit around the table, speak to people on the screen? If not smell, what is absent that makes that less valuable?
SRINIVASAN: A really good question. I’ve thought about this a lot, actually. There’s two technologies that are obviously going to converge. The first are these Anybots or Double Robotics-type things, which are basically like an iPad on wheels, which you can remotely control like a video game, roll it around, et cetera. The second are the Boston Dynamics–style humanoid robots, which can do grasping and other kinds of things. They’ve got a humanoid form factor.
Eventually, those two technologies plus a VR bodysuit are going to be combined, such that you can, maybe in 10 years, script a robot from around the world. You’ll be able to basically see through the eyes of the robot, and you can actually actuate it and move the arms through the VR bodysuit.
COWEN: I’ve done this giving a talk, by the way. I don’t know how it was for the audience, but it does, in some literal sense, work.
SRINIVASAN: Right. The thing is that the robotics have to be better. The bandwidth has to be better. All that stuff has to be better. And when it does get better, an interesting thing happens. Your immigration policy becomes your firewall.
That is to say, someone can “immigrate” to your country over an internet connection, connect to a robot, animate it, and start walking it around. Now your ability to interdict those internet connections is your immigration policy.
That has massive implications for all kinds of things. Among other things, you wouldn’t have to travel thousands of miles to get a job as a day laborer. You could literally comparison shop between jurisdictions. Even migrant work–type labor could eventually become software, which is really amazing because then they could get a higher wage. Potentially, they could always go to the best market.
There’s going to be some really interesting things that happen with telepresence in the medium term. We’re not there yet, about 5 to 10 years.
COWEN: There’s some evidence, say, from soccer that peripheral vision is an important determinant of whether you’re a good soccer player or a great soccer player — how much you take in around you that is not obviously the thing you’re tasked with doing.
Maybe one problem with telepresence, it’s simply hard to test or try out other people’s peripheral vision because exactly they’re focused on looking at you. You don’t see how they take in the surprises because they’re insulated from those because they’re pointed at you in some way.
SRINIVASAN: I think that is solvable though. A lot of the VR work nowadays is focused on field of view and peripheral vision and making it feel more immersive. You just need the right kind of cameras on the robot and the right latency and whatnot. That’s potentially solvable.
But you’re right: It does need to become more immersive. Eventually, though, it may cross over and become something where people prefer that than travel because they can go to Egypt and they can go to Japan and they can go to the UK all within the same day without any jet lag. They’ll literally move at the speed of light.
On how tech is changing compulsory education
COWEN: Here’s a very general question. Right now, children, especially in the Bay Area, they grow up with so much tech. It surrounds them all the time — iPad, smartphone, apps, whatever it may be. They learn how to use it very young.
There are costs and benefits to this. Overall, I would say I’m optimistic about it. But just in general, what is the implicit ideological commitment or value judgment we’re making by just allowing this to proceed? How do you think about that decision?
SRINIVASAN: The state has six hours of compulsory education per day. What’s interesting is the network is getting a lot of those hours and taking them away from the state because the kids are plugged into their laptops and iPads and iPhones and self-educating during the period they’re supposed to be having compulsory state education.
There’s an interesting thing that’s happening there, where you’re having kids plugged into a different network, not the state-approved network, during their nascent years. They’re plugged into the internet and their own sub-communities and subcultures. So the state has a lot less control over kids as they’re coming up.
The state has six hours of compulsory education per day. What’s interesting is the network is getting a lot of those hours and taking them away from the state because the kids are plugged into their laptops and iPads and iPhones and self-educating during the period they’re supposed to be having compulsory state education. There’s an interesting thing that’s happening there, where you’re having kids plugged into a different network, not the state-approved network, during their nascent years. They’re plugged into the internet and their own sub-communities and subcultures. So the state has a lot less control over kids as they’re coming up.
If you go all the way back to how Bismarck instituted public education and whatnot, it was really to raise children who were “obedient to the state,” who were patriots or however you want to call it.
Because that acculturation is happening less and less because kids are getting plugged into folks who have different ideologies at a much younger age, I think that’s got a lot of interesting medium-term consequences.
COWEN: Some parents would say there’s a potential maybe addiction risk, if that’s the right word, or maybe an attention risk. Then on the plus side, there’s a lot more information, more freedom of thought. If you’re an unusual child, you can connect with others like you more readily. How much do you think we know now those are positive tradeoffs?
SRINIVASAN: I think that, on balance, it’s more upside, more downside, like everything technology is doing. That’s my one-liner for the future: more upside, more downside.
In this context, of course, downsides are obvious. You have kids who are less able to make eye contact because they’re staring at a screen all day. Literally maybe, their eye physically can’t focus at medium distance as much anymore.
You’ve got kids who — it’s harder for them to focus on books and to do deep reading because they’re constantly distracted by all these notifications. The downsides are obvious.
The upsides, of course, are that you have access to the Library of Alexandria for free any time. If you’re really good at math or computer science or have any interest like that, you can actually find other people.
You’re not forced into this one-size-fits-all 12 years of quasi-jail that modern American public schools are. Instead you can self-educate and self-advance and level up. So on balance, I think it’s probably better. It’s probably positive on net. Though we want to figure out some way to ameliorate those downsides.
On exit and voice
COWEN: You’ve given a very famous talk, or several talks actually, on the ideas of exit and voice in the global community — the idea that your ability to leave a government can induce liberty and wealth-creating policies.
If we have a theory where exit is important as a source of liberty, how do we explain that liberty seems to be declining in all of the world’s leaders, or most of them?
SRINIVASAN: Now, I think, the whole axis is tilting back the other way where, because they were civilizationally vaccinated against the lure of communism due to the 20th century, the Chinese, the Russians, the Eastern Europeans have a civilizational memory of why that was a bad idea.
So they’ve reacted against it by establishing these states, which are not necessarily all totally free-market capitalist in the best ways but certainly are not what you’d call hardcore communist. They’re not mass-murdering people and seizing their land anymore. That affects society in many different ways. Basically, people have this near-term 30-years-ago-memory of “Hey, that was actually a bad idea to try to move in that direction politically.”
COWEN: Right now, are you more bullish on large political units or small political units?
SRINIVASAN: Small countries in general, with one major exception, which is China. I think in general, the countries that are going to win the future will be Estonia, Singapore, Israel — very software-savvy small countries where their leaders are basically almost like software CEOs.
What I mean by that is, for example, Estonia’s president, Toomas Ilves, literally a computer scientist. Lee Hsien Loong, who is Lee Kuan Yew’s son, was the wrangler at Cambridge, very smart guy. Netanyahu is an MIT graduate. Those are three examples of folks who are pretty software savvy as leaders of their country.
I think in general, the countries that are going to win the future will be Estonia, Singapore, Israel — very software-savvy small countries where their leaders are basically almost like software CEOs.
China’s also, as a large country, extremely software savvy. The Great Firewall is an instrument of policy. That’s something that they actually think about as a first-class thing.
Even Putin actually has made recent comments about the blockchain, showing it’s on his radar. He’s maybe U-turning Russia to make that a national priority for them.
I’d say, in general, I’m more bullish on small countries. There may be a few large ones that are software savvy enough to compete.
COWEN: When it comes to cybersecurity, quite arguably the form of warfare of the future or even the present, is it the small or large countries that have the comparative advantage in either attack or defense?
SRINIVASAN: That’s interesting. Israel has some great cybersecurity people. So does Estonia. So does Singapore. So does China — gigantic. It has a lot. I’m not sure you can initially correlate that to country size. I would say, however, there’s a bunch of these small countries that punch above their weight from a software standpoint.
COWEN: Peter Thiel has suggested that, with the onset of artificial intelligence, this will encourage at least authoritarianism of the negative sort, or possibly even totalitarianism. Do you agree?
SRINIVASAN: I thought Peter had an insightful juxtaposition or contrast between blockchain’s decentralizing aspects and AI’s centralizing aspects. I do think that it’s a little bit complicated by the fact that there’s recent breakthroughs or developments in AI which allow for calculations of the coefficients locally on your phones.
Without getting super technical, up to this point, typically the way you’ve done machine learning is you’ve centralized all the data, have huge data tables, datasets at your central servers, and then run the analysis.
But there’s potentially ways to do it where you basically do a MapReduce out to all of your phones, and you calculate it locally. The data and everything stays locally. This is just a coordinator, but it’s not storing it. Google’s published some research on that. We’ll see how big that becomes in the years to come.
On the future of identity and privacy
COWEN: But say I’m in Shenzhen right now, as I understand it, there are several intersections where if I jaywalk, they will do facial surveillance on me, very likely identify me, find me automatically. There’s a very strong element of control there. In China, if you have a bad social credit score, maybe you can’t ride on the train or take a bus.
Why isn’t the future simply easier control of people in this way, which would produce some more local public good but nonetheless be a kind of horrible dystopia that most of us wouldn’t want to live in?
SRINIVASAN: I think China may well be vectoring in that direction.
The thing about it is, what they’re almost moving AI towards is almost like the all-seeing, all-knowing god of myth. Some of the stuff they’re doing with face recognition is taking all the surveillance footage they have and then turning it into something which can take somebody’s face, pull it out of that, pull it into a database, and then track a human being across the entirety of their life essentially.
That takes it from, “Hey, I need to look at the surveillance camera and recognize this person” into something where the computer is recognizing them for me. It just completely levels up the threat level of what a surveillance camera is.
What do I think that results in? It may result in a very low-crime society. It may be something that certainly does have lots of obvious negative aspects, and I probably wouldn’t want to live in a country like that. But it might be something that a lot of people nevertheless choose to live there.
There’s folks who choose to live in places that you or I or folks in the West would not choose to live in. Maybe they think, “Hey, the tradeoffs are worth it.” I wouldn’t. Probably most westerners wouldn’t, but I’m not necessarily saying that that’s going to cause their civilization to fail.
A lot of people thought, for example, that the Great Firewall in the early 2000s — because of the restriction of freedom of speech — would cause the Chinese internet to collapse, and it’d be a disaster, and they wouldn’t be able to get innovation and so on. They managed to get very far with the controlled internet.
I have a little bit of a, for lack of a better term, modesty on this, where as bad as I think that future is, (a) I’m not sure it’s within our power to prevent it, and (b) I’m not sure the Chinese people would want the West to prevent it, or at least not all of them. That’s where I am on that.
COWEN: If I call up, say, Apple customer service and I need to prove I’m Tyler Cowen, one thing they might do is ask me for the last four digits of my Social Security number. This seems absurd. For one thing, that kind of information can be quite readily available.
The future of personal identity — proving you are who you claim to be — say, 10, 15 years from now, what will that look like? How will we do it? Is it a retina scan? Is it fingerprint? Is it facial recognition? How will you prove to the world you’re you?
SRINIVASAN: Good question. I think it’s going to vary depending on the context. For example, retina scans and biometrics are good in that they’re unique to you. They’re bad in that there’s no password reset, for example. If password reset is a core feature of your security system, you may not be able to use that with biometric authentication.
I think we’re going to have a bunch of different options, from email and phone number auth, username auth, biometrics, et cetera, where it’s just going to fit the particular use case.
COWEN: The exit option — this is debated quite a bit now with Facebook, with Google. Some people say, “Well, Facebook, you don’t like it, you don’t have to use it. You can just exit.” How persuasive do you find this argument?
SRINIVASAN: It’s certainly possible to delete your Facebook account. It is possible, with probably some effort, to ban all Facebook.com domains from your local DNS. It’s harder to exit Google, but also possible. It’s certainly possible to do it if you went to China because all that stuff is banned and you can have a modern life.
I understand that the counter to that counterargument is the sheer number of trackers, the ubiquity of it, and it’s hard to live a normal life without having some interaction with those companies. I’m mixed on that in the sense of I personally know how to exit Facebook or Google or whatever if I wanted to do so.
I recognize it’s too hard for other people to do so right now. But I think that the right approach is to build exits and build alternatives such that folks can exit from those things and they don’t feel constrained by them.
COWEN: Let’s turn to cryptocurrency for a bit, an area you know something about. Which of the arguments against the future viability of crypto assets do you find most persuasive, granted that you’re not agreeing with them? What’s the best criticism? If someone says, “Oh, cryptocurrencies in 15 years, it won’t be a thing,” what’s the best argument for that point of view?
SRINIVASAN: That the decentralization model actually doesn’t work and that it’s actually possible to compromise the underlying blockchains. Last year, I had an article called “Quantifying Decentralization,” which put forth a metric for how decentralized a given coin is.
The idea is that it could be decentralized in terms of how many miners you’d need to compromise to control it or how many developers or how many exchanges or what have you. And the minimum number of entities that you needed to compromise was the degree of decentralization of the system.
The utility of that — I’m not saying it’s a perfect metric; it’s imperfect like all metrics — but the utility of that is that you could start to rank order coins by their at least purported level of decentralization. I think insofar as these blockchains fail, it would be because the decentralization models were not robust enough to protect against the attacks they’re supposed to protect against.
COWEN: Will libertarians eventually turn on Bitcoin as some commentators have predicted?
SRINIVASAN: There’s certainly been a civil war within Bitcoin, the Bitcoin Cash versus Bitcoin Forks. I don’t think either of them would call themselves less libertarian than the other. It’s just an intra-denominational fight.
COWEN: How will we limit the energy-intensive costs of bitcoin mining?
SRINIVASAN: That’s a good question. There are coins that are being developed that have more energy-efficient proof of works or proof of work alternatives, like Bram Cohen’s Chia. Bram Cohen invented BitTorrent. He’s a legit guy, and we’ll see what he can do. Proof-of-stake and delegated-proof-of-stake are a variance that are basically like shareholder votes.
COWEN: But you’re reintroducing third-party involvement in some way. If you’re asking how decentralized is the system, someone has to verify who has stake or share or whatever.
SRINIVASAN: Right. I would say I don’t have a good answer to that other than, obviously, no exponential can go to the moon. It’s going to top out and logistic out at some point.
With that said, I will offer one observation, which is that people will compare Bitcoin’s power consumption to, let’s say, PayPal’s and say, “Oh, PayPal just has a few servers and Bitcoin is using this gigantic set of server farms. Look how wasteful it is to send a transaction.”
But I don’t think that’s an apposite comparison because what Bitcoin is, and then more generally Ethereum and all the blockchain stuff is, is basically a system of property rights and contract law. It’s more comparable to, let’s say, Delaware or even the US government, at least certain functions, than it is to a central server sending payments back and forth.
The blockchain records who has the property. I think that’s something where, if you’d make that comparison, even all of Bitcoin’s energy consumption is cheaper than the Brink’s trucks and the US military and the police system and so on required to protect property rights in the American context.
The blockchain records who has the property. I think that’s something where, if you’d make that comparison, even all of Bitcoin’s energy consumption is cheaper than the Brink’s trucks and the US military and the police system and so on required to protect property rights in the American context.
COWEN: Right now, I pay financial fees to my mutual funds, to Merrill Lynch, all over. Anytime I save money, I’m paying a fee to someone. Which of those fees will go away?
SRINIVASAN: Good question. Maybe all of them. If you think about a major business model for the US Postal Service, prior to the introduction of email, with stamps — now most of the mail that you send doesn’t have any stamps associated with it. Instead, you had a completely different business model, which is you show ads in email.
COWEN: But I can do those accounts electronically already for some time, but still I’m paying fees. They’ve gone down a bit, but the financial sector seems to take up about two percent of national wealth, and it has for many decades, roughly.
Why should we think that will go away? We’re paying for some kind of service; maybe it’s not clear what it is. What in tech is going to do that more cheaply? And what will be done more cheaply for us?
SRINIVASAN: Because you can set up a new payment rail and contract execution system in your dorm room. That’s new because it provides fundamental competition at the lowest level. Now, these new payment rails are volatile. They’ve got all kinds of issues associated with them.
But you can actually, root and branch, exit the system in a way that you couldn’t have done before, and I think that’s going to give more fundamental competition to fees. The closest analogy is, with the internet, you have a new mechanism for transmitting information which lets you route around the Postal Service and NBC and the televisions and the movies and the record companies and whatnot.
In the same way, you’ve got this new rail or new set of rails that lets you route around everybody. I think that’s going to structurally reduce fees over the years to come.
COWEN: I can right now buy crypto assets through third-party intermediaries, and they, too, will charge me fees. It just feels, as a customer, a lot like the fees I pay to my other intermediaries. It’s true I have the option of doing it myself, but for most people, that’s very, very costly to figure out how to secure a wallet and how crypto assets work in the first place.
They would rather pay the fee, basically, of two percent of their wealth. So why aren’t we just reintroducing intermediaries at some other stage of the game, and we’ll end up paying fees just like we used to?
SRINIVASAN: I think there will be a business model of some kind. What you call fees may shift to another part of the system.
As an example, Germany recently repatriated several tons of gold from the United States — their central bank did. That was something that took three or four years and cost so many millions of dollars. In theory, the same amount of money could have been moved in Bitcoin for a fraction of that cost. There’s already certain contexts where we’ve reduced fees by several orders of magnitude for very large movements of money.
Then what’s going to happen is, these new business models that arise are not going to have exactly the same fee structure as the old ones. There will be a business model. They’ll still continue to make money, but they’re going to tax what is scarce, and that may not be necessarily the wire transfer itself. But they may charge some money management fees or other kinds of things like that.
The most important thing, from my standpoint, is the level of choice you’re going to have is going to be dramatically increased because the blockchain makes it much easier to startup a competitor and raise up your own shingle as a new financial outlet.
COWEN: A question about tokens. Here’s a statement. Tell me if you think it’s true, false, or uncertain. The role of the ICO is to aggregate dispersed information about a company. ICO, initial coin offering.
SRINIVASAN: I wouldn’t necessarily agree with that. I think we’re in very, very early days on tokens and so on. One useful thing that I heard from someone on a panel was, she believed that tokens were going to become an integral part of every company’s capital structure over the next 10 years. Her argument went as follows.
She said, “OK, what is debt? When a company issues debt, it’s a claim on the future cash flows of the company. What is equity? When a company issues equity, it’s a claim on future liquidations of the company. What are tokens, or what are network tokens? They’re a claim on the company’s future digital asset value or digital network.”
In the same way, the interplay of equity and debt is something where a competent CFO knows how to trade off between them. Not every company is suitable for debt financing. If you don’t have cash flows, you’re not suitable for it. Not every company is suitable for equity financing. If you’re not going to do 10X, if you’re a mom-and-pop grocery store, equity financing may not be for you.
Finally, with tokens or network tokens, unless you have a network effect, token financing may not be for you. On the third hand, some of these companies are suitable for those means of financing. I think, in addition to network tokens, you’re going to see tokens that are basically debt tokens, equity tokens, and other kinds of financializations. All the regulations need to get worked out.
We’re at very, very early days on the whole token thing. And an ICO bundles a lot of different concepts together that are going to get unbundled and separated out in the years to come.
COWEN: Let’s say I’m an old style company. I’m Nordstrom or I’m Macy’s. Those companies — for a long time, they’ve issued gift certificates, which in a way raises money, brings in capital, maybe a small amount. We’re familiar with how this works.
If we wanted to, we could trade those gift certificates amongst ourselves. We could even trade them on blockchains, but that probably wouldn’t be seen as a major innovation. Maybe the price of the gift certificate would tell you something about how good is the merchandise at Nordstrom or Macy’s.
ICOs are more than that, somehow. What’s the element they bring to the problem that you don’t get just by trading, say, gift certificates of Nordstrom or Macy’s?
SRINIVASAN: That’s a good question.
COWEN: Or subway tokens, right?
SRINIVASAN: Right. I would distinguish, just terminologically, between the ICO itself and the token because you can have a token without an ICO and potentially vice versa. But let’s say tokens. Why is a token different from just a gift certificate, for example?
The first thing is, you have, in theory at least, total custody that say you own the private keys. You can move it to somebody else without any third party’s intervention. Whereas probably a Macy’s or anybody who’s issuing these gift certificates has some restrictions in terms of transfer or what have you.
COWEN: But you give them as gifts. That’s why they’re gift certificates. They have numbers.
SRINIVASAN: That’s right. The other aspect is, once you can do something programmatically, a lot of things become unlocked with it. You can have these tokens. You can have millions of them. You can have programs consume them. You can have them bought and traded programmatically. You can use them as inputs for API calls.
There’s things that you can do with them on a computer that you can’t do with a physical gift certificate. The agility of the value dramatically increases, and I think that’s a very important aspect. That’s not just like a physical gift certificate.
COWEN: Say we peer 20 years into the future. We consider a person who doesn’t have a college degree and is at the median income level for the United States. Twenty years from now, they’ll be saying, “Oh, because of tokens, I can do X.” Maybe they don’t understand how it all fits together, but they do understand they can do X, and this is meaningful to them. What’s the most likely candidate for X?
SRINIVASAN: One thing I’ve thought about, and I think this is a 20-year thing, maybe something like a personal token, which is to say, if you have some future . . . Let’s say you’ve done well in something that shows you have some promise. Rather than, let’s say, taking out a student loan, which is one way of mortgaging your future, you might issue a personal token.
That would be something where you say, “Hey, look, I’m going to sell about 20 percent of my future earnings for token-based financing. Here’s my résumé and here’s all my stats. Can you finance me and help me achieve my dreams?”
That’s something where anybody in the world can chip in. A guy from India, a guy from Japan, a guy from Kenya, Brazil can help finance that. Maybe they’ll lose money, maybe they’ll make money, but it opens up financing in a way that’s never been possible before at a very small level, a very individual level.
I think that’s a 20-year vision. I don’t think that’s a 5-year vision. Maybe a 10-year vision.
But in the same way that we put up a website for any old thing, you put out a tweet for any old thing. A communication, a broadcast to millions of people is now such a low-threshold activity. If financing or the financialization of something also becomes such a low-threshold activity, it’s going to become a lot more common.
On things under- and overrated
COWEN: In all of these conversations, or most of them, there’s a segment in the middle called “overrated versus underrated.” Feel free to pass on any one of these. I’ll toss out a few things or names and you respond. Ramanujan, the Indian mathematician, overrated or underrated?
SRINIVASAN: I think he’s underrated because he’s not yet world famous. He’s famous in India.
COWEN: Why is he important to you?
SRINIVASAN: He’s from my father’s hometown. He’s India’s Einstein. I think, also, he’s very spiritually important in the sense that he represents the rags-to-riches story. Even though he didn’t get rich, he came from the middle of nowhere. He was very poor, destitute, but he was able to level himself up and really make a mark in the world.
If you want to say what drives me is the power of the internet to do something like that. To find all these people who are in the middle of nowhere, who have been left behind by the 20th century, and give them the information, and then, with the blockchain and the financial tools, to lift themselves up and lift their communities up.
COWEN: Indian food in Silicon Valley, overrated or underrated?
SRINIVASAN: Exactly right rated, I guess. It’s fine. I have no opinion on it.
COWEN: Why? What will they do that we haven’t thought of?
Or like the example I mentioned where you can teleport into a robot and then control that, Skype into a robot and control that on other side of the world. That’s going to be something where maybe you’re going to have it in drone mode so it walks to the destination. You’ll be asleep and then you wake up and it’s at the destination.
Drones are going to be a very big deal. There’s this interesting movie called Surrogates, which actually talks about what a really big drone/telepresence future would look like. People never leave their homes because, instead, they just Skype into a really good-looking drone/telepresent version of themselves, and they walk around in that.
If they’re hit by a car, it doesn’t matter because they can just rejuvenate and create a new one. I think drones are very, very underrated in terms of what they’re going to do.
COWEN: The novelist and poet Vikram Seth?
SRINIVASAN: I’m a fan of his. I remember reading his book . . . I think it was A Suitable Boy.
COWEN: That’s my favorite.
SRINIVASAN: Actually, no, it was Golden Gate. That’s what it was. That was a book that was in the form of a poem. And at the end, he had some really moving, melancholy lines about “Oh, we’re all going to die someday.” I was like, “Ugh, this is horrible. Can we not work on that?” That’s what got me really deeply looking at telomeres and life extension and all that type of stuff. That was 18 years ago.
I still think that’s going to happen. There’s folks here in the Valley who are working on that. There’s a lot of prerequisites, but there’s been a lot of progress for the last 10, 15 years.
COWEN: Not counting anyone you’re working with now, who would be three quite underrated or undiscovered independent thinkers in tech today?
SRINIVASAN: I would say, at least from the American context, I’d say the Chinese internet entrepreneurs. I know that’s a broad category but . . .
COWEN: Sure, that’s fine.
SRINIVASAN: Jack Ma, Pony Ma. The Chinese internet is a lot more innovative than, let’s say, the broad mass of the US has realized. One way of looking at this in the show Silicon Valley the opening title has only one Chinese internet company there, like Alibaba. It doesn’t really have the rest of them. They’re not as actively featured.
I’m not critiquing the show. I’m just saying that it hasn’t impinged on the American consciousness, the degree of innovation that’s happening over there, whereas in China, they’re very hyper-aware of what’s happening in the US. So it’s kind of asymmetric. So I do think the Chinese internet is very underrated.
Second group of, let’s say, thinkers, doers. I’d say that the global English-speaking population — again, this is not one individual but what I think of as underrated.
There’s many more English speakers outside the United States than there are within if you include India, if you include Nigeria, if you include all these countries that have relatively large English-speaking minorities.
Now that those folks all have smartphones, it’s only a matter of time before they’re really heavily involved in US discussions because what the internet is doing is bringing people together, based on interest, from different locations, cryptocurrency being a great example of that, where you’ve got people from around the world who are participating in this thing.
There’s going to be more things that are like that. Now, you’re going to get tens, hundreds of millions of English-speaking people, maybe, participating in the US internet. That’s going to be very different and very new for people.
COWEN: With respect to China, if you think about the best Chinese tech companies in terms of management, do you think they have any advantages over the Bay Area?
SRINIVASAN: I don’t have as much firsthand knowledge, so I’m going to give . . .
COWEN: Do they have quicker product cycles, as some people have alleged?
SRINIVASAN: What a lot of folks say — and I’m not sure the extent to which that’s true because I haven’t experienced it myself — what people will say is that the average Chinese tech worker works harder than the American tech worker. There’s the 12-hours a day, 6 days a week kind of thing.
The big difference is, the Chinese that I’ve met, at least, really feel like they are trying to show that China can make it on the world stage. They have a hunger to prove something that isn’t as much there . . . For the Americans it’s more of an individual thing as opposed to a civilizational thing.
You saw this to some extent on display at the Beijing Olympics. Ten years ago, China wanted to show, “Boom, we’re not a backwards country, and we’ve really arrived.” That’s an aspect which, to a greater or lesser extent, is present in a lot of the Chinese entrepreneurs I’ve worked with, and I think that’s much less present among the Americans.
On the ideal India trip
COWEN: Let’s say you were asked to design for a highly intelligent and curious friend, who had never been to India before, the ideal India trip. How would you think through what that might look like?
SRINIVASAN: It depends on whether they want the authentic India or the tourist India. The authentic India has got a lot going for it. The tourist India will be much closer to the western experience.
Given those parameters, I’d say you either stick to the most manicured city aspects and probably the south, which is a little bit more attuned to western sensibilities in terms of just HDI and whatnot.
Or if they’re up for it, you could give them the authentic experience and just take a train to any place in India. It depends on their preferences.
COWEN: What’s the one place they might not think of going where you would say, “Oh, you must go here. This is more interesting than the guidebooks will tell you”?
SRINIVASAN: That’s a good question. I’m not much of a tourist. I’m not sure I could say for something like that. Maybe Goa. I think Goa is a unique place within India that is a cultural mixture because it’s got the Portuguese influence, Indian influence, et cetera. I couldn’t say on that.
COWEN: What do we Americans in the tech world least understand about Nigeria?
SRINIVASAN: It’s English-speaking. It has a very large number of people. There’s now starting to be some software stuff coming out of there, like Andela. I think there’s going to be a lot of those folks who work in what I call the new digital blue-collar jobs, which are Mechanical Turk–type jobs.
We’re doing some of that with Earn.com, where you have labeling of images: is this a cat, this a dog, or determining if there are two clusters or one cluster in an image. That’s what I think of as the new digital blue collar, like worldwide machine learning–type stuff. You’re going to have a lot of those from Nigeria, Brazil, India, places like that.
COWEN: How do we do a worldwide talent search for the really smart kids? Is that a problem that can be solved by tech, or is it more about local mentoring and coaching?
SRINIVASAN: I definitely think that is a problem that can be solved in nontrivial part by tech, though you’d want to use some local mentors to help localize it. The way I would do it is — this is actually something I’ve thought a lot about — essentially use a MOOC and take the highest scorers.
To some extent, that’s already being done. For example, even five years ago or six years ago, there’s an article in the NYT called “The Boy Genius of Ulan Bator” about a kid in Mongolia who was in his early teens and outscored a lot of kids at MIT on this circuits class. So that’s already being done.
What hasn’t happened yet, though, is to take those kids who are scoring well and then plug them into the global economy so they can contribute and get their just dividends. That’s the kind of thing that I’m excited about that maybe the blockchain will play a partial role in.
On the Balaji Srinivasan production function
COWEN: For our final section, I have some questions about what I call the Balaji Srinivasan production function — that is, how you got to be you. If I understand correctly, you grew up on Long Island, right?
COWEN: Both of your parents were physicians. How would you say your high school experiences formed your worldview in a significant way?
SRINIVASAN: It’s not something that I spend too many cycles on, but it is something where it’s like, “Oh, wow.” I’m envious of kids today who have Stack Overflow and GitHub and a really functional Google and whatnot. I come from the generation that was just before the internet really started to get useful.
High school was kind of stultifying in the sense that you just couldn’t learn anywhere near as fast. I feel like my education really only commenced in 2001 when Google started to get good enough that I could really start leveling up on the internet and just self-educating and drinking it in.
Insofar as my upbringing shaped me, I would say it was just something that taught me, “Wow, I never want to send my kids to an American public school. I’ll probably crowdfund a tutor or do something like that so that they can actually learn something rather than being in jail for twelve years.”
I have this one-liner which says, “Life in the United States begins with a 12-year mandatory minimum, the Schoolag Archipelago.”
COWEN: You have several advanced degrees from Stanford. One is in chemical engineering, which is slightly unusual. One of my readers wrote in to me this question: “Why get a degree in chemical engineering?”
SRINIVASAN: Oh, there’s a method to the madness there. I had a lot of credits because, since high school was really boring, I took a lot of AP classes, and I independent-studied a lot of AP classes. I had a ton of credits coming into Stanford, and nothing to spend them on.
I wanted to get into genomics, and the Human Genome Project was going. What I wanted was the major that would expose me the maximum to genomics while also having the maximum amount of math because I didn’t want to just memorize a lot of stuff. So chemical engineering was that, where it had a lot of biology, but it presented it probably in the most structured form.
The other thing was that, at least at Stanford at the time, chemical engineering and electrical engineering had the reputation as the hardest engineering majors, and so I wanted to do it just because it was hard.
COWEN: What’s the most important thing you learned from your brother Ramji?
SRINIVASAN: My brother Ramji is very good at execution. He is incredibly disciplined, very, very smart, and he just approaches problems in a different way than I do. What Ramji contributed is a skill or a mode of thinking that I call “list, rank, iterate.” It’s kind of a meta-algorithm, which is almost trivial to describe but a very useful way to attack unstructured problems.
Basically, what it means is, let’s say you have a problem. “How do we increase sales?” or “How do we raise funds?” or “How do we do X?” or “How do we do Y?” Often what you can do is, you can start by making a list — for example, a list of prospective doctors to sell to or a list of prospective VCs to go and fundraise from.
Then impose a ranking function on them, which is, “Hey, which zip code is this doctor in? Are they likely to prescribe this test?” et cetera, or “Has this VC invested in companies like ours before?”
Then you just iterate brutally through this. The key thing is you put a limit. You say, “All right, I’m going to do 150 of these. If I don’t get any hits on them, then I’m going to try a new strategy, a new list.” This concept of list, rank, iterate is a great way of structuring unstructured problems. That’s something that I owe Ramji.
COWEN: Your upbringing from Long Island, how has that shaped who you are?
SRINIVASAN: How has that shaped who I am? Well, I actually don’t think about it that much. I don’t know if you’ve seen those movies where someone’s frozen in a block of ice, and then they’re thawed out, and then they experience the world. That’s kind of what my life has been like post coming to Stanford.
COWEN: Long Island was the ice or the world?
SRINIVASAN: The ice, yeah, exactly. It was kind of like everything in life moved in a very slow motion until around 2001 or thereabouts. Then things sped up a lot.
I don’t know if you’ve seen those movies where someone’s frozen in a block of ice, and then they’re thawed out, and then they experience the world. That’s kind of what my life has been like post coming to Stanford. It was kind of like everything in life moved in a very slow motion until around 2001 or thereabouts. Then things sped up a lot.
COWEN: What’s the best theoretical framework for thinking about optimal family size — how many children you should have?
SRINIVASAN: Interesting. I plan on having as many as possible. The reason for that is, I like children and I come from large families and so on. It’s a very individual decision. I think Bryan Caplan has an interesting book arguing for large family sizes.
In some ways, the folks who have large families are going to win the argument simply because they’re going to have more kids and those kids will have more kids, et cetera. It’s not really [laughs] even a logical argument. It’s an evolutionary argument. Darwin’s going to have the last word.
COWEN: And it will keep the world somewhat religious, on average . . .
SRINIVASAN: Maybe, yes.
COWEN: . . . because religious families tend to have more children.
COWEN: If you think about the different obstacles you’ve encountered to get to where you are, there are a lot of clichés about what matters — be smart, be conscientious, five-factor personality tests — from Mark, I’m sure you’ve heard.
COWEN: But outside of that framework, if there’s some source or influence or insight you would bring to bear on the question of how to be successful that you feel is underemphasized in the Bay Area, what would it be?
SRINIVASAN: That’s an interesting question. Underemphasized in the Bay Area?
COWEN: Correct, because everything else was this block of ice on Long Island, right?
COWEN: Nothing is emphasized except the cold.
SRINIVASAN: Sure. What is underemphasized? The degree to which you really can and should set out to set your own path, rather than looking at TechCrunch and Twitter and whatnot. There’s a certain model of person who is successful because they go, and they look at the competition, and they align around the competition. They’re just a fast follower and so on.
But it’s very hard to tweet or TechCrunch your way to innovation. I really do think you have to tune out a lot of what the Valley is thinking and then innovate. For example, I can’t believe Satoshi got the idea for Bitcoin by reading TechCrunch, that just completely out-of-left-field idea that had no validation from it.
So I’d say tuning out the Valley is often the best way to make a huge impact, and then have the Valley recognize that you’re a real innovator. Some of the biggest innovations like SpaceX — no one was telling Elon to do that. There was no feed about how space was the next big thing. There was no conferences on space or whatever. It was not thought of as a thing that entrepreneurs could conceivably do.
What I almost look for, what I’d say is, take everything that people are talking about and then the rest of human civilization, and look for the stuff that technology has not moved into yet. That’s maybe where the opportunity is because it’s underexploited.
COWEN: If you want to tune out the Valley to some extent, do you ever think of not living here? And if you couldn’t live here or chose not to, where else would you live?
SRINIVASAN: I’ve been wanting to get out of here for a long time.
COWEN: To where?
COWEN: Why Singapore?
SRINIVASAN: It’s the optimum place for at least my personal preference factor because it’s hot. One thing that’s little known, by the way, is a lot of the South Asians in Silicon Valley have low-level vitamin D deficiency because the place isn’t really as sunny as you might think, and we’re spending all this time in front of our computer. It’s a small reason, but it’s a long-term health thing.
In addition to that, Singapore is friendly to capitalism. It’s got a lot of skilled immigrants there. You can start a tech company really easily. It’s safe. If you’re raising kids, it’s good. It’s just a good place. I like it a lot. It balances a bunch of factors. There’s other places that have a lot of those attributes, but they’re cold, like Switzerland can get cold. So Singapore is probably number one on my list.
COWEN: Final question: when all is said and done, what do you hope your legacy will be?
SRINIVASAN: I definitely want to work on the blockchain for a while and then eventually get back into biology. I’ve got an idea for something in the genomic/quantified self space. Hopefully, that makes an impact. Let’s see in a few years.
COWEN: Balaji, thank you very much.
SRINIVASAN: Thank you, Tyler.