The net is under assault by both too-powerful companies and ill-equipped countries. We users are getting left behind, and there is real risk the net’s global unity could be compromised. How can the world better oversee and regulate the net?

The following transcript has been lightly edited and condensed for ease of reading.


Kirkpatrick: So this next session is very close to my heart, as any of you who read Techonomy’s journalism will know. And the only reason Nick Thompson is moderating it instead is because he is so goddamn good at moderating and he also is deeply immersed in this issue. I remember one time, like 10 years ago, seeing Nick interview Peter Thiel at the Council on Foreign Relations and it was like, “Wow, I want to be that good.” So he is great. So thank you for Nick, who is the editor of “Wired” for being here.

And the other three—in general, we generally don’t like panels with this much testosterone, but it just happens that these three guys really know their stuff. So maybe you should all come up.

Veni Markovski is the ICANN representative at the United Nations and a long-time friend of mine. He’s Bulgarian. An entrepreneur who had entrepreneurial success in his country and kind of came to the States and he’s been doing all kinds of great stuff. Really knows what’s happening on the Internet.

Scott Malcomson is one of our next two panelists who both worked in the Obama administration. Scott worked in the State Department. Dipayan worked in the White House. Scott is a writer—what’s that book you wrote, Scott?

Malcomson: “Splinternet.”

Kirkpatrick: “Splinternet.” He wrote a book called “Splinternet” that I—the name is sufficient definition.

Dipayan has been not only in the White House, but worked for Facebook for about two years in privacy policy and he’s not there anymore. And he’s now at the Shorenstein Center on Media, Politics, and Public Policy at the Kennedy School. And he and I have gotten to know each other partly through some mutual work we’re doing around these issues.

I’m very excited to hear them try to explain how bad it is and whether it could ever be less bad. So Nick, take it away.

Thompson: All right. Thank you very much, David. Thank you for that very kind introduction. This is one of the issues that I worry about the most and it is the most interesting. The world is splitting apart into different internets, right? There may be a Chinese internet, an American internet, a European internet, an Indian internet. It’s dividing. And not only is it splitting apart, the relationship between the big tech companies that make up the Internet and the users is not so great, as is alluded to in the little joke about Dipayan’s prior work at Facebook.

So we have a world splitting apart, the big powers and civilians not really getting along so well, the government’s not sure what to do—the government’s not even understanding the tech, and we have to figure out what to do. So, we have 25 minutes and at the end of the 25 minutes we will have solved this problem. Let’s get cracking.

I’m going to start with you, Scott. So you at least introduced the word “Splinternet” to me—I think it’s your word, and you think we’ve gone from a splinternet to a winternet. So let’s explain what the splinternet is and how it evolved into a winternet. And—Let’s begin there.

Malcomson: Yes. So thank you. I didn’t invent that term, but I wish I had and if this room accepts that I invented it, that’s the next best thing. Winternet I think was original.

I’m just going to say quickly I’ve been working lately with a kind of a three-part periodization for this topic. The first period being the Internet as we kind of still kind of think of it. You’ve got your basic plumbing, which Veni handles. You’ve got your, you know, your protocol layer, and then you’ve got content. And what happened seven, eight years ago or so China was thinking very seriously about having a separate route and part of the political bargain was that this three-part structure would include content within the sphere of the state. In other words, the state could affect content more or less as it chose. So that, to me, is period number one.

Period two, which I think we’re in now and have been for a little bit, I call the “Appernet.” I’ve got a little business in neologisms. So the Appernet is if you think of WeChat, you think of your experience of the Internet as being either entirely or largely mediated through apps and that’s how the state can increase its control. And it leads to different business models as well, which I’m hoping you might talk about.

And then the third stage, which I think we’re moving into, is, once again, more neologisms, I’m calling “Game of Clouds,” which is basically where data localization and 5G and EDGE networks all kind of meet. You need cloud computing, in order to provide all these benefits of low latency, tends to be physically associated with a particular part of land, right? And it needs to be nearby and states led, once again, by China with Internet sovereignty are trying to locate their clouds—essentially have national clouds. And I think that’s going to further transform the way that we use and experience the Internet.

Thompson: Do you think the issue of state control began at level one and will continue to, sort of, to level three into this Game of Clouds period?

Malcomson: Yes, absolutely.

Thompson: Okay. So Dipayan, let’s go to you. So who’s the good guy in Game of Clouds? Who do we actually—who doesn’t have dragon—

Malcomson: Take us to the last episode.

Thompson: —flaming their users?

Ghosh: Well, I wish I could tell you all who’s going to sit on the throne at the end of the series, but I think really what we—what we’re living in now with the Internet today is a commercial zone that really lacks any sort of regulatory standards. And when we think about kind of the political, economic tradition of this country, it’s always to put the markets first. And that’s what we’ve done in the case of the Internet. It was a public domain. Over the past 20 years, in particular, it’s become tremendously commercialized. And this has resulted in a few good ideas really burgeoning, taking advantage of the network effect, and becoming the dominant players in the modern information and communication medium that we all use today.

And I’m talking specifically about a few big Internet companies: Facebook, Amazon, and Google among them. And I think, Nick, what we are seeing now is a very consistent business model across the consumer internet really focused on three things in particular. Creation of tremendously compelling platforms that are borderline addictive, some psychologists say, and which are tremendously engaging for individual consumers and close out competition over the Internet. Second, the uninhibited collection of data on individuals through those services, all toward the end of creating behavioral profiles on those users. And third, the creation and refinement of algorithms that are tremendously sophisticated and opaque that essential curate our feeds and target ads at us.

And if you look at—if you look at Facebook, Google, Amazon, or any of these other consumer internet companies that are dialogical with consumers, it is that business model that really elevates these companies and allows them to have the profit margins that they currently have. And I think we can draw a direct line from that business model to all the negative externalities that we’re seeing today from disinformation to hate speech.

Thompson: Okay. But do you guys agree or disagree? So Scott just said it’s kind of a concern about the way the governments are thinking about the Internet and you just laid out a whole bunch of ways where it’s structural problems with the big tech companies that we need to worry about. So for those of us who just want to preserve the Internet, which of you guys is right? Dipayan, do you agree with Scott or do you disagree with Scott?

Also, please tweet this session and use those addictive platforms to get the message out. Instagram it away and let’s go from there.

Do you guys agree or disagree, Dipayan?

Ghosh: Well, I think I absolutely agree that we want a certain kind of Internet. We want certain kinds of standards and the standards that China, Russia, and similar countries are perpetuating and trying to establish for their own domestic regimes are not the style that we want in this country. In fact, for those countries that are on the fence like India, like Brazil in regard to the way that they manage and govern the Internet going forward in their domestic situation, I would hope these countries will follow an approach similar to the approach followed in the United States.

So it can—all I’m trying to say is that both of these problems can be pretty significant. And I think that this problem of tremendous capitalistic overreach can happen in either situation, especially one that favors the markets without any sort of consideration for the end consumer.

Thompson: So we have capitalist overreach, we have government in the United States not doing enough, government in China doing too much, and we’re now going to watch how the rest of the world plays out.

Veni, you were called a plumber about three minutes ago. Tell us how to fix the plumbing here.

Markovski: Well, I mean first of all, I just need to make sure that, you know, I don’t speak for ICANN. I work at ICANN, but only the president and CEO can speak for the organization. But I do spend most of my time at the United Nations and other intergovernmental organizations which are dealing with issues related to the Internet. And like this year, we’re going to see a lot of that dialog and conversations happening at the UN about cybersecurity.

So we, as an organization, care about the domain names, IP addresses, and protocol parameters. These are issues which most of the users of the world don’t know. They may know that the domain name exists, but they really don’t understand what’s the understructure, how it translates into digits, and what are these digits doing, et cetera, et cetera. Because it takes milliseconds for the machines to figure it out. And we actually—that’s the beauty of the Internet as we know it, which is one Internet united around the world. That wherever you are, you use the same protocols, you use the same domain names, you use the same databases around the world to make sure that when you type something, you go wherever you want to go.

So in terms of fixing it? I mean, it’s already been fixed many years ago. What we are trying to do is we are updating it constantly. We are increasing the security of the Domain Name System, like introducing DNSSEC, which is a protocol for security—securing the DNS. And we are making sure that this model works around the world.

And now of course our model is very unique because ICANN is an international organization but not a treaty organization. And we have like 400 people around the world—

Thompson: You guys don’t have an army, right?

Markovski: We don’t, as far as I know.


Thompson: But are you worried about the underlying structure of the Internet splitting off as this tension between the United States and China and India grows?

Markovski: I think Scott mentioned about the Chinese, few years ago, talking about having their own root servers. The reality is that the root servers are the ones that exchange this information around the world. An internet is a network of networks, so we should not think about the Internet as one network. Every organization, even this facility here where we are, you know, has their own network and it’s part of the Internet and they can create their own rules about accessing it or their own limits. Like you may have noticed that the bandwidth here is not very high, which is amazing, you know. Why is that? Somebody has decided it.

So every network decides their own policies and then once they connect to the rest of them with the same protocols and they use TCP/IP, that’s it. They are part of the Internet.

So in some ways, we continue to evolve this process. We continue to evolve the policies that relate to domain names and IP address allocation. We encourage IPv6, the new version of the Internet protocol addresses, but we don’t do that in—in like top-bottom way. We are a very bottom-up organization and that’s why I said it’s a unique multi-stakeholder model where everyone—government, businesses, civil society, academia, individuals—everyone is on an equal footing with everyone.

Thompson: So that makes me feel a little bit better. You guys scared the hell out of me and now we’re talking about the underlying root structure and it sounds pretty good.

So Scott, why don’t you scare us all again.


So explain a worst plausible scenario for the Internet dividing over the next couple years that you see happening. What does that really mean?

Malcomson: Well, there’s so many to choose from, but so I mean, just very quickly. What Veni’s describing is what I call the plumbing, which sounds like not that complimentary, but it’s absolutely essential like plumbing’s essential. The Appernet and the Game of Clouds and everything, that’s on top of the existing Internet. It’s not really changing that underlying structure. It’s changing the experiences that you can have and it’s changing the balance between private and governmental control of what you do over the Internet, which is sort of a separate thing.

In terms of scary scenarios? I mean, the one that I’m most worried about is the—you know, basically, I think the reason a lot of states, particularly China, but others, have developed such a keen interest in these topics lately is because they see security threats. Not necessarily political opposition security threats, but an inability to control the technology of—hard technology of security. You know, whether the missile actually hits the target and that kind of thing.

And so that state fear has triggered a lot of this move towards an assertion of sovereignty over the technical space. And so my fear is that that trend will continue and my hope is that, at some point, maybe through looking a little bit at the history of the Internet, we can see that it is possible to have a non-militarized Internet. But you have to think about how to do that. It’s not at the plumbing level. It’s at other levels.

But that, to me, is the single most important topic because having been in government and been in the security world, once you get onto that escalator, it’s super hard to get off. And governments are on the escalator now. They see a threat to their actual physical safety through the kinds of network systems—

Thompson: So what happens at the top of the escalator? Like what does that mean?

Malcomson: Well—

Thompson: They’re on the escalator. What happens on the second floor?

Malcomson: So, okay. So here’s an example. If you’re all familiar with the F-35? None of this is unclassified. Joint Strike Fighter. Very popular new weapon of the Air Force spreading around the world. It’s essentially a sensor platform. We have a lot of allies who are buying these from the United States. In about 20—15 or 20 years, if everything goes as expected, about 70% of the F-35s in Asia will be non-American. They will have bought them, but they won’t be American planes. But we’ll be relying on that network to act and react in hot security situations.

So just very quickly, the example is those things, those platforms have so many sensors and they react so rapidly that what the US military is thinking now is that it needs to push, as we say in the commercial space, push the decision making further towards the edge. And what that means is that you’re starting to take decision making about pulling a trigger or not away from the top of the chain of command because the speed at the edge of the network is so rapid.

And so that’s the kind of scenario that’s on the second floor. If we let the network aspect become so overwhelmed by the security agenda—it’s not just the US, it’s just our example. You end up creating literally a hair-trigger situation, so I think we need to figure out how to back away from that.

Thompson: Dipayan, you’re nodding along with this, but this is kind of a different take from your take. Tell me why you’re agreeing with Scott here, because it doesn’t seem like something you’d agree with.

Ghosh: No, no, I agree that we want open norms. We want a governance of the Internet that the US, traditionally, has tried to pursue. But I also think that the Splinternet, as Scott puts it, is one problem and another problem, which may be of even greater magnitude, is the ongoing commercialism of the Internet which has really disturbed, I think, the social contract in this country. If we think about these companies that have really dominated the consumer internet, every single silo that we think of, whether it’s social media or video sharing or search or internet-based text messaging—each of these sectors within the consumer internet is dominated by one company in the United States market.

And so you could make the case that in each of these silos, market silos, there’s a monopoly and that they’re actually extracting monopoly rents against the consumer and potentially damaging, in the longer term, our media ecosystem. And I think this is exactly—we can draw a line from that situation to things like the disinformation problem and the spread of hate speech.

Thompson: But is the solution to that to make the US government more aggressive in claiming control over the Internet, possibly leading to some of the problems that Scott is talking about in the militarization of the Internet? Or is the solution for the US government to be less controlling and try to support competitors? What is the solution here?

Ghosh: I think there can be a happy medium. I think we don’t need to make the case that the US government should go all the way toward how China is trying to control its domestic situation, but rather—

Thompson: Excellent, excellent. We’ll agree there.

Malcomson: Stop short of that.

Ghosh: Yes, somewhere short of that.

Thompson: You still prefer our political system to Xi Jinping’s?

Ghosh: Yes. Yes.

Thompson: Excellent. Okay.

Ghosh: But what I’d suggest is that instead of having a radically open free market for Facebook to run wild, that we set some borders around that business model. And if the business model, again, is focused on, let’s say, anticompetitive behaviors at the expense of potential rivals and unchecked data collection and opaque and potentially damaging algorithms, then you could make the case then that what we need to do to renegotiate the social contract is push new regulation around competition and privacy and transparency to really give the consumer some power back.

Thompson: Okay. Veni, let’s go to your domain but not so much your ICANN, the lower levels. Let’s talk about you as, you know, a man who advised the government of Bulgaria on IT policy. How likely is it that we’re going to get the countries of the world and the governments of the world agreeing on sort of general policies for how the Internet should be run and that will be able to prevent the Splinternet?

Markovski: Well, how it should be run? Again, it depends where, exactly, on which level. On the level of the domain names, the government’s members of the governmental advisory committee of ICANN—

Thompson: On domain names, we seem good.

Markovski: 177 of them, they participate in our meetings. They’re very actively involved.

Now, on the policies that every government decides what to do in their country—I mean, you mentioned Bulgaria and I’m always proud to say that for the last 20 years, since ‘99, this is the only country that has said that even though governments around the world in the telecom laws controlled the names, numbering, and addresses, Bulgaria is the only country which by law says “Except Internet addresses and Internet names.” So that government, back in ‘99, decided this.

And for the people who live in New York or anywhere in the US, actually, that may be a revelation. The result of this lack of regulation of any kind in the Internet is that with 7 million population, so less than New York City, there are about 2,000 internet service providers.

Thompson: Which is about 1,998 more than in New York City.

Markovski: Exactly. So the result is high speeds. Like, I moved 15 years ago to the US and there was like—I was having 1 gigabit connection back then in Bulgaria for which I was paying $10 dollars. It’s the same now and we cannot think of this in the US now because everything is so expensive and the bandwidth is still low.

So I think the policies—but that model was not taken, even though the Bulgarian government was sharing it widely around the world. That model was not taken by other countries because the culture is different, the businesses are different, the territories is, you know, you cannot do it on a big scale.

Thompson: And the governments fought for. I mean, the ISP—well anyway, keep going.

Markovski: And no. So my point is we—I don’t—the only place where governments are now discussing somehow at least the cybersecurity aspects of their policies is the United Nations. And that’s where we come very helpful as an organization because we provide educational outreach to the United Nations, to the diplomats that work there, both in New York and in Geneva. Because we want them to know how the Internet actually works. I’ve been bringing here experts from the Internet Architecture Board, the Internet Society, ICANN CTO, and others—you know, board members—to explain how the Internet works. Because when they go behind closed doors in these negotiations, we want at least a few people to know how the Internet works, because otherwise, they may take a decision which will impact everyone without actually knowing whether it’s possible to achieve their goal using technical or other means. So that’s a very challenging task because these are foreign policy—I mean, foreign ministries’ diplomats. So they’re not like the telecom people who come to the ICANN meetings, from the telecom ministries. So it’s a very challenging too. And they rotate every three years, you know. There is somebody new. And I think we don’t pay enough attention in the US about what’s happening at the UN. People kind of ignore it, but a lot is happening and this year, in particular, there will be two different groups to discuss cybersecurity, which has never happened before within the General Assembly. So how are they going to discuss that parallel to each other? What kind of reports they will issue? Then the Secretary General has a high-level panel on digital corporation also coming with some reports next month. It’s so fascinating for the people who are inside. Not so for the people outside because it doesn’t make the news, you know. There is no blood on the floor or anything like that.

Thompson: Okay. That’s because the social network platforms don’t incentivize news that doesn’t have blood on the floor. We’ll put that on them too.

You made me feel good again. We just have a minute and 45 seconds left, so let’s—it’s 2019. What is the most—the thing you would be most optimistic about if it happened this year? As we think about these problems, like, what is the—we got, you know, it’s a train. They’re rolling towards each other. They’re going to crash. I don’t know what the metaphor is. But what is the best thing that could happen that would make you feel good?

Markovski: Well, the best thing will be the trains actually are on different rails so they don’t collide.

Thompson: Yes, there we go.

Markovski: That would be the best. But I also think the trains are actually going in the same direction. It’s just that they’re going with different speeds.

Thompson: The Bulgarian one’s going faster.

Markovski: Well, I’m not going to—I don’t want to speak so much. But no, the trains are with different speeds because there are different people on the trains, different engines and, you know, the technical world is somehow working in one way, the policy world is working in a different way. So if we manage to find the common ground between those two—bring them together—that’s the way. And that’s actually the ICANN’s model.

Thompson: All right. Scott, what do you hope happens in the next year?

Malcomson: Well, you mentioned Brazil and India earlier. I’m really interested in how some of the larger markets are reacting to this US-China so-called great power conflict. And I think it’s kind of—I hope, since we’re trafficking in hope now—I hope that for those larger markets, that they’ll start to develop more of an independent path.

There’s a really interesting company in Nigeria called MainOne which lays cables and builds data centers and everything, and they’re doing a lot of the cloud work for West Africa. Just the other day, Huawei got its contract at the Belt and Road Conference for Kenya. I prefer the MainOne way, given a choice, and I think that’s—I hope, rather—it’s unfamiliar to me to say that, but I hope that these other markets will take lessons from that, and I think some of them are that they really need to carve out an independent space. That they can’t let the large platforms—for example, Facebook in India—kind of shape their future for them. And maybe the scariness of the current situation will push that forward.

Thompson: Dipayan, final optimistic thoughts?

Ghosh: Well, I think over the next year, I think the biggest thing that can happen is—well, we’re seeing a tsunami come for Internet companies, for Silicon Valley. A tsunami of regulation from India, from Japan, from South Korea, from Europe, and I just hope that the industry can come together with legislators and policy makers in this country and really set some norms around privacy and competition, in particular, and transparency, political transparency, especially, to really start a new discussion. And if we can make progress in that area, it’s going to have big impacts around the world, I think.

Thompson: All right. Well, we made progress on this panel. Thank you very much for listening. Thank you, excellent panelists.