 Tonight, at this first night, we have someone here who's going to introduce us to how to inscript, resist, and subverse in the governance of the Internet infrastructure. I hope, at least. I'm quite interested in what Nielsen-Uwe has to tell us here, because we might not be afraid. You told us that we don't have to be afraid about your talk. So keep your weapons down, keep your phones open, and I give you the word. Thank you. Give a warm welcome to Niels. Let's go. I am super excited to be here today with you on MCH, and I'm wired up like Beyonce, so let's hope I can give you the stuff that you are looking for tonight. So thanks so much for being here on your first night and chose not to be dancing, but to actually think and talk about the governance of the Internet infrastructure. I got a little test here on the screen, and that says that the Internet architecture social technical imaginary guides, facilitates, and legitimates the Internet infrastructure, but actually the infrastructural norms serve to evaluate and instruct the governance of the Internet. Right now, you don't understand what this means. By the end of the talk, you will. So let's run the test again by the end, and then we'll see where we are and if I've succeeded. I'm Niels De Nouver. I'm a postdoctoral researcher at the University of Amsterdam, and we'll talk to you about my PhD thesis, and it's very cool that I'm talking about this here, because this work actually started at home. And that's where I met some of the people that I'm still doing the research now with. So that's something I'm very excited about. The work continued at Shah, and now I'm presenting it to you at MCH, so that is great. I got five of the copies of my book for you, so if you actually like it, come and pick it up. Who wants a book? Yes? Can someone distribute it? Thank you. So then you've got a way to read it in book. The PDF is in the description, so you can read it that way as well. And at the end of the talk, I've got a fun announcement for the future. So I look at communication infrastructures. And I do that because I think that changes to the globalizing world are being written not in the language of law and diplomacy, but rather in the language of infrastructure. And if infrastructure is the language of the changes, then standards are its grammar. And that's why I look also at the standard setting of the internet. And when I talk about the internet infrastructure and its governance, I explicitly look at the interconnection and physical layer of the internet. The physical layer are all the antennas, satellites, cables, and routers. And the interconnection layers are actually what make it into one network. And this happens in institutions of three and four letters, such as the IETF, ICAN, RERs, IEEE, 3GPP, and ITU. But this is not a new thing. And infrastructures have been around for the longest time. Around in the 19th century, the British had a telegraph line that covered all of the world. And that was important because it changed the British Empire from a colonial to an imperial power. It allowed the British to have control directly from London over an empire that vastly expanded topological means, topographical means, and they controlled it through topological means, through a network. Communication networks are always a proxy for power and are intrinsically linked with militarism and colonialism. The telegraph, as I said, changed colonialism to imperialism by establishing centers of power. So networks established new centers of power. And in war, for instance, in the First World War, the first thing that happened was that the Allies cut the connections of the German telegraph lines. Now in Ukraine, we see again that infrastructure is an object of a text. However, there is a difference right now because all parties seek to make the network function not only for them but still connected into this global network. This says something about the meaning of the Internet. We now live in late capitalism, where countries were ready to shut off networks of capital such as Swift. But they were not ready to impose sanctions on the Internet infrastructure. So what this means, that Internet and communication infrastructures are foundational for how the whole world is run. So it's probably very important that we look at how these networks are governed because that also says something about how the world works. And infrastructures are always relational. They are not just the physical things, the things that we can kick, but there are also the standards, the agreements, the labors and the institution that make them function. And when we talk about networks, we should not think about that global network as one thing because something can be a global network and be very different in different localities. So we should listen to people on different parts of the world, such as the great thinker in Ukraine, Svetlana Matfajenko, who is working and writing in Ukraine right now. Please follow her work. But it also means we should think about people who have other practices and languages. For instance, Ramzi Nasser once wrote a programming language in Arabic and he called it engineering performance art. And I asked him, why Ramzi, do you call this engineering performance art? Why is it not just engineering? Well, he said, because it will never work, because the internet structurally rejects everything that's not in a Latin language. That means that the earliest application of the internet, or one of the first, email, works perfectly for many of us, but if your name is in a non-Latin language, for instance, if you're Chinese, then you cannot have an email address with your own name in your own script. That means that that global internet is very different for other people in different areas with different languages and it can also make people at home, but it can also exclude people and it does this all the time. So submarine cables now connect the whole world, but if we actually look at how it's ordered, it provides a particularly ordering of power, extraction, and control that reproduce repressive power structures as they exist and reinforce power structures as they have existed over the centuries. So in this first part of the talk, I'm going to cover a part that looks at specifically the internet engineering task force. So the internet engineering task force is one of the oldest bodies that produce the internet. And you might know it from RFCs, who of you has ever read an RFC, right? So then you know that if you start that an RFC is an abbreviation for a request for comments, but if you start commenting once it's an RFC, you're actually too late because then it's already a foundational document, but previously it's called an internet draft. And when it's an internet draft, people from the internet engineering task force discuss how it works over mailing lists. This is a photo of the internet engineering task force in Toronto, Canada. And this was actually right before the meeting, during the meeting where they decided that the internet engineering task force could not have a meeting in India because the infrastructure there was too bad. And what then happened was that the power went off. And what was so funny that the meeting just continued in the glow of all the screens of the laptops and no one actually gave even a small hiccup. So this also shows ironically that it's the people that make the infrastructure work on what they expected to work and how they expect to do business. So to do this, it's important to understand how this work is guided. And for that, I borrow a term from sociology, from science and technology studies. It's called a socio-technical imaginary. And a socio-technical imaginary is a vision of how something should work that brings people together to create something. And the electronic superhighway is an example of such a thing. So what is so interesting about the IETF is that it produces these RFCs and they've done it since the end of the 60s in the same format. And all the debates happen on mailing lists, which is great for researchers such as me because it allows us to do computational analysis of all their conversations but also to analyze all the archives of all the discussions of how the internet works, what are the rules of the road. So I did a quantitative analysis of all RFCs, a qualitative analysis of 50 RFCs, qualitative and quantitative analysis of mailing lists, a total of 35 gigabytes and almost 2 million emails of 955 email lists and participant observation during four years and 11 meetings of the IETF. So what I thought I would find there was many different socio-technical imaginaries of what people thought the internet would be, what it would be for them. But what I found is actually there were three things that they all said that the internet was, three foundational principles. The first one was the end-to-end principle. And the end-to-end principle says that intelligence in the network is at the edges, the network only exists for datagram transport that it has low complexity and higher robustness. But at the same time, when you look at an RFC 1958, this end-to-end principle is not just a technical principle. Because read with me here. The community believes that the goal is connectivity, the tool is the internet protocol, and intelligence is end-to-end rather than hidden in the network. And while doing participant observation, you find many different ways where values and ideals are stored. One of them is on t-shirts. Many of you will be wearing t-shirts which communicate ideas and identities. And here it also shows that this is also inherently political. And this is then also shown again in another RFC 3935, and it says the internet isn't value-neutral and neither is the IETF. We want the internet to be useful for communities that share our commitment to openness and fairness. We embrace technical concepts such as decentralized control, edge user empowerment, and sharing of resources because those concepts resonate with the core values of the IETF community. These concepts have little to do with the technology that's possible and much to do with the technology we choose to create. This sounds more like the manifesto of a group of anarcho-syndicalists than a standards body of multinational corporations that run the internet. Right? So that's a funny thing to find. So let's see how that works out. So there's the second principle of permissionless innovation, and I cannot reiterate enough, almost every interviewee said these were the main things. So permissionless innovation says that there should be no barriers to the deployment of new protocols. There should be no need to negotiate with entities in the middle of the network. And this is probably a response to the telco era, where in the telco era, intelligence was in the network and the end nodes were dumb. The end nodes were dumb and the network was smart. But now, in the internet, it was turned around and the network was made dumb just to transport packets and the endpoints were smart. But it was also a response, arguably, to the accessible use policy of the ARPAnet and the NSFnet that said that no commercial traffic was allowed on the internet. So up to the early 90s, no commercial traffic was allowed on the internet. So if anyone ever tells you that the internet is a product of companies and capitalism, it's not true. It was publicly funded, made at universities and with government money. But in the 90s, this changed. The third principle was that of openness. And that has three parts. Network openness, that any endpoints on the internet should be reachable for any other endpoints. And that new endpoints can be added. That it should consist of open standards that are voluntary and freely accessible and of open governance that's transparent, open participation and have open archives. Exactly the principles that allows me to research it today. And that then results in very interesting configurations, such as this, the unofficial motto of the IETF, which is we reject kings and presidents and voting. We believe in rough consensus and running code. And this was a very explicit reference to the International Telecommunications Union that has governed previous networks, such as the Telegraph, TV and Spectrum, that does voting and that is governed by countries. No, the IETF would do so based on rough consensus and running codes. Does anyone know how rough consensus is established in the IETF? It is indeed humming. So if we want to decide about the length of an IP header, and once you'd say 128, say 265, and I would be a working group chair, if you believe the header should be 128 bits, please hum now. You've got the concept, right? So that is quite amazing that this is how the internet is governed and how the decision-making is designed, right? So the work of Sandra Brahman shows that explicit discussions about rights and freedoms, as we've seen in these RFCs, as well as social impact of technologies, have featured in RFCs since their beginnings. So now if we take a step back earlier, I already gave it away, if we need to pinpoint the start of the internet, what event would it be? What is this? Sputnik, the launch of Sputnik Satellite in October 1957 by the Soviet Union. It emitted what Time Magazine would call those chilling beeps. Beep, beep, beep. This device that was exactly my weight, and it was shot into the atmosphere, and that was actually there, was a symbol of the technological and scientific advancement of the Soviet Union over all civilizations on the planet, and that instilled enormous fear in other countries. This led the US to invest in a network that could resist a nuclear strike, and that caused the distributed nature of the internet. So in the end of the 80s and the early 90s, what happened? The fall of the Berlin Wall, right? So that's when the US government seeded direct control over the ARPAnet, the NSFnet, the ESnet, and it led to the establishment of commercial internet exchanges and to the formal institutionalization of the internet governance regime that we know now, such as the Internet Engineering Task Force, the Internet Society and regional internet registries. But at one of these first meetings of the Internet Engineering Task Force, a German statistician raised his hand. Somehow, these statisticians are almost always German, I don't know why. But he said, if we continue handing out IP addresses at this rate, within three years, we cannot add new devices to the network anymore. And that would be a great problem because then the internet could not live up to the concept of openness anymore, because new endpoints could not be added to the network. So what happened? The introduction of network address translation. This meant that routers would have a public IP address, but that there were private IP addresses that would be shared over several devices. What happened as well? There were increasing security considerations. The internet was no longer comprised of trusted actors, so firewalls were instated. And there was a perceived need from network operators to differentiate business models and provide quality of service, caching, and prioritization of services. So a lot of networks started to do a lot of bad things to the network. They were not just solving IP, but they were adding control inside the network. And this rise of the middle box added functionality to the network, not at the edges of the network, but in the network itself. And this led, already in the beginning, to an ossification of the network. So in this introduction of directionality of not boxes, which caused not all endpoints to be able to connect to all other endpoints directly, introduced a directionality which created two classes, clients and servers, users and producers. And that made the importance of servers way more important than that of other devices. This device has way more capacity than any web server had for the longest time on the internet. But still, we cannot simply host a website on this because it's not reachable because of carrier-grade nuts. And this created a new affordance structure in the internet architecture. And an example of this is TLS 1.3. So you're all familiar with transport-layer security that's allowing for the S in HTTPS, the most used secured encryption algorithm on the internet. But in the development of TLS 1.3, they came across a big problem. People were experimenting with TLS 1.3, but it wouldn't work on the network. Why not? Because all these middle boxes were expecting only TLS 1.2. They did not know what TLS 1.3 was. So they would simply drop the packets of TLS 1.3. So TLS 1.3, to be deployed, now needs to lie to all the routers in the middle by saying, actually, I'm TLS 1.2, but then when it reaches the endpoint test, oh, sorry, I'm actually TLS 1.3, but I had to lie to the network. This means that if we want to design the network of the future, it needs to look like the past. This is the result of middle boxes and ossification of the network. But TLS 1.3 was actually still quite a success, and this cannot be set for the stream control transmission protocol. The stream control transmission protocol was supposed to be a more efficient replacement for TCP. It had multiple streams, multiple transmission paths, no head of blocking, and it's actually described in 39 RFCs, and it worked perfectly in the lab, but was blocked by almost all nets. And this means it never reliably worked on the internet. It worked in almost, it works on pretty much all telco systems, but it never works on the internet. So the first RFC of SETP was published in April, 2002, and the last RFC was published in November, 2017. And people indeed blamed the network for this. But someone swifed in to solve the problem through the rise of Qwik, the Qwik UDP internet protocol, which is a stream-based protocol similar to SETP, but developed by Google. And Google could develop this because it could do A-B testing between the biggest browser in the world, Grome, and almost server's everywhere in the world, Google's content distribution networks. And probably you have been taken part in the testing of Qwik without knowing it, because you were using Chrome in many different environments and thereby passing information back to Google to see how and whether it works and which versions of Qwik actually worked. And if it didn't, it would simply fall back to TCP. So Qwik is actually a pretty great thing because it includes TLS 1.3, so encryption, it does congestion control, loss recovery, and it's built on UDP, pretty much the way I'm presenting. I'm sending you way too much information all the time. So what it does do with this encryption after Snowden, of course, got a lot of interest, but it does not only encrypt your information, it also ensures that only you and Google have that information. And all control over data streams is taken from the hands of the network operators so they cannot inspect the traffic anymore or route it based on stuff they get from the data analysis. So what we see in TCP is SIN, SINAC. AC, if you do it with TLS, SIN, SINAC, then we negotiate keys and TLS control, and so you have a lot of back and forths. In Qwik, this would go way quicker. So Qwik was a win, right? So all well, that's end well. Well, maybe not, because only a large transnational corporation such as Google that had control over the largest browser in the world and the largest CDN network could gather that information to develop a new protocol and actually roll it out. Qwik tooling is not readily available for everyone yet, and Qwik deployment will arguably strengthen consolidation of Google and these large parties because other people have less access and possibility to shape and change information streams. Also, it is much harder for privacy researchers, ironically, to understand what Google is doing. So all the research done on cookies is not as easily done now because network researchers cannot look at the data streams and the practices of Google because the line between you and Google is encrypted at all times. And the network operators, one can think, one you know, are not pleased because they also have less control over their data streams. So the former chair of the ITF told me that you now, if you want to deploy something, you need to play in some of the operators or vendors earning models in order to get something deployed. So that is quite different from the imaginary at the beginning where it said that everyone could deploy everything on the internet all the time, right? And if we then look at the authors and affiliation of the authors of the RFCs, we see that actually a relatively small number of companies are setting the rules for the development of the internet architecture and infrastructure. So these myths, these social technical imaginaries, might be important for what they reveal, including a genuine desire for community and democracy, but also for what they conceal, including the growing concentration of communication power in a handful of transnational media businesses. So the social technical architecture imaginary and its self-regulatory governance model have not been able to safeguard freedom and equality of researchers, small companies, or individuals to innovate on the internet protocol level and permissionless innovation for the purpose of retaining openness has undermined itself and the end-to-end principle. Corporate interests have become a first-order considerations for protocols to be adopted and implemented, and political conceptions of the architectural imaginary are fading into the background. The importance and size of the internet architecture has only grown as we now see that everyone wants to keep everything connected all the time, and with it is societal implications. But the societal implications are not structurally considered in the standardization of the internet. So some considerations have taken up, such as the mentioning of security in RFCs. So here in 1990, it became obligatory to have security considerations in RFCs. So they are mentioned in all RFCs. Privacy is increasingly mentioned. But if you look at the length of these security considerations, there's a lot of copy-paste, and they might not be as thorough as we want them to be. So to think then about other conceptualizations, together with Corinne Kaath and other people, we developed human rights protocol considerations to develop ways that protocol developers could think of how their work impacts human rights. So to show you how this was actually taken up in the IETF, I show you the first human rights protocol considerations in the first RFC that came out after this, namely RFC 8492. That says, the most fundamental of human rights is the right to protect oneself. The right to keep and bear arms is an example of this right. Implementations of TLS passwords can be used as arms kept and born to defend oneself against all manner of attackers, criminals, governments, lawyers, et cetera. Therefore, TLS password is a powerful tool in the promotion and defense of universal human rights. So you see, there are very different understandings of human rights in the community. So now let's continue to the second part, where we look at norm conflicts in the governance of transnational and distributed infrastructures, the case of internet routing. So in this next part of my research, I try to understand how norm setting and norm conflict work in distributed and complex transnational infrastructures, such as internet routing. So you know about internet routing, right? You know that the internet is a network of networks. These networks are autonomous systems. Autonomous systems have an autonomous system number. These are assigned by regional internet registries and regional internet registries also hand out IP blocks to these autonomous systems. And between them, they route internet traffic. And that happens through the border gateway protocol that does announcements who has control over which IP blocks and filtering happens based on that and is secured through RPKI. Sorry if I go a bit fast. Sometimes I'm a fallen over letter box. But if you have questions, we have time for that. So you see all these different networks. They all have different numbers. And they route traffic between them if they don't have a direct connection to a network where they want the connection of the computer with this IP address wants to reach the computer with this IP address. It asks the computer of this network, asks, hey, do you have this network? No, but I'm connected to that network. And through that, packets get routed all over the globe. And if one connection gets cut, it finds its way through other ways through that network. So this is a way of the way all these networks are connected. But that seems to be a very technical and not very value-based approach. But some really important thinkers of early internet architects already set in an important paper, different stakeholders that are part of the internet you have different interests that may be adverse. Accommodating this tussle is crucial to the evolution of the network's technical architecture. But currently, routing decisions are made based on a mixture of efficiency, trust, and economics. And this is facilitated by the RIRs I earlier mentioned. So I thought, let's design an experiment if we could do routing based on other principles. So what principles could these be? These should be relevant, of course. They should be applicable. And these should be existing frameworks. I didn't want to come up with a new existing new framework. And these frameworks should be transnational, such as the internet. So what are the frameworks that you think that I could use for this? Of course, you're all beloved GDPR. So if a network says, hey, we respect the GDPR and you as a user only want to route your traffic through networks that respect the GDPR, it would be cool if networks say, hey, I respect GDPR, right? I thought that could be useful. Or even if networks could say, hey, I respect human rights and particularly the United Nations guiding principles on business and human rights that made it possible for also to understand if companies respect human rights. So for that, I proposed an introduction of two objects in the RIPE database, namely the ASSET GDPR, which would say that the members of this set declared to be compliant with the general data protection regulation of the European Union or the ASUNGP, which would say that the members of this set declared to have adopted and implemented the United Nations guiding principles on business and human rights. So I announced this. I tried to get networks to sign up to this so that we would actually get to route over only human rights respecting networks. Wouldn't that be great? People didn't agree. No one signed up to these ASSETs. Why? Seems like a good idea, right? So I argued this is because of infrastructural normativity. And the interview, I based this on the interviews I did with the people that responded to this. So norms are the common values of society and the collective expectations of proper behavior of actors with a given identity. And they have four aspects, namely identity, behavior, propriety and collective expectations. So from that I found that there is an infrastructural norm of voluntary interconnection that network operators are expected to produce interconnection based on voluntary bottom-up collaboration and trust. And they believe that the network facilitates and engages in interconnection through the BGP protocol that in this case, RIPE coordinates and facilitates interconnection through the RIPE database. Everyone has control over their own networks and interconnection is based on incentive structures and enlightened self-interest. And the reason given to me why GDPR and human rights should not be a basis for routing were based on technology saying, well, the routers do not really allow for that. Why will we prioritize implementing that in software? For an institutional response, the database is not meant for that. This is not what we intend to do. We don't do politics. We are here to be neutral and impartial. This is not how it's done based on identity, but also economy. There is no economic incentive to do this. No other company asks for this. No user asks for this. Why should we do this? So what norms get resisted? So everything that complicates or hampers interconnection. Laws such as the GDPR, but also ethical frameworks such as the UNGP, jurisdictions such as Schengen Routing, which was also rejected, but also sanctions and security measures. And this was also something that was reflected in a recent call when the Ukrainian government asked for the blockade of Russian IP addresses. RIPE swiftly responded, the RIPE NCC board quickly responded that the executive board of RIPE NCC is committed to taking all lawful steps to ensure that the RIPE NCC can provide undisrupted service to all members across our service region and the global internet community. The RIPE NCC will publicly document all its efforts to ensure that the registry is not negatively affected by laws, regulations, or political developments. So the RIPE NCC is committed to take all lawful steps available to ensure that the Russian military keeps on being active on the internet. You could call it impartial, as RIPE NCC does, or not so neutral indeed. So the voluntary interconnection norm instructs the community of network operators to create more interconnection between and among networks and resist any norm or value that could hamper that. So the internet governance is guided by internet architecture, social technical imaginary, anchored in architectural values and process values, such as openness and transparency. And the social technical internet architecture imaginary serves to guide and facilitate the process of norm development and legitimizes the current institutional ordering, even if they are concretely subverted in practice. Remember, only Google could actually do that. And infrastructural norms exist to instruct new norm development and evaluate norms that are being introduced. And the introduction of norms is only successful if the candidate norm serves the interest of these significantly represented groups. And the new norms are not in conflict with the overarching infrastructural norm of voluntary interconnection. And then by now, you should be getting this. So the internet governance regime, I'd argue, undermines its own authority, scope, and legitimacy because it resists norms that are in the interest of nation states and individuals. We need to rethink the network ideologies that underpin the foundation of our information society. The internet infrastructure get much less attention and platforms, algorithms, et cetera, even though it impacts everything that happens on top. We need new network ideologies. Right now, we might be at a new Sputnik moment. Come think with us. And for that, we are establishing the new critical infrastructure lab to research power contestation in the global digital infrastructures. And we're doing that together with Maxigas and Fika. And so if you wanna think with us about this, we will be thinking and working on this for the coming years. So please come talk with us after this, ask questions, and we'll hope to be studying this with you. Come find us today, tomorrow, or any given day. And thanks so much. Happy hacking. Thank you, Nils. Thank you both as well for collaborating here. It was a fantastic speech. I expect there are lots of questions here. We have a couple of minutes. Are there questions from the internet? Let's start there. No, no one. Then someone here in the hall. Okay, please start, take a microphone there. There's another microphone in the back, so if someone wants to queue. Hello. You said that the introduction of network address translation kind of led to this split into clients and servers. You think maybe IPv6 could undo this in a way? Thank you. So for the longest time, this is what we hoped, right? But if we now see, we see different things happening with the deployment of IPv6. And the first thing was to be observed that there were carrier grade knots all over IPv6. But at the same time, we also see that censorship measures are not working as effectively on IPv6 as IPv4. So for an end user, there might be some advantages, but if we look at structural ways in which network are managed, unfortunately, we see that a lot of these behaviors from IPv4 roll over into IPv6. And that's where we see that it's really not just the technology, but also the cultures and identities and incentives of the engineers and the companies that are working on this to not have completely separate stacks, but to manage IPv6 in the same way people have managed IPv4. So unfortunately, the way it looks now, not yet. Yes, the next question, please. Hi. So you showed the participation at ITF, which is heavily skewed towards large companies. Do you want to invite the hackers to actually join ITF and the regional registry communities to change what's happening there? This is a question of conscience. And I think I don't have the answer. So I think there is not one way to the future. There are many different ways. But I think that it's very hard, as you've all known, within your company, within your own cultural groups, within your own social groups, it's very complex and hard to change a group's culture. And it's very hard to change embedded interests. So I definitely think we should participate, regain control, regain knowledge and understand how things work by participating in these bodies. But at the same time, we should also think of alternative economies and alternative ideologies on which global communication networks should be run. So yes and no. So I do think that we have a lot to learn from these institutions. And don't get me wrong. I don't think these institutions are bad, but they have developed in a particular way to run the networks based on particular cultures and particular incentives. When I introduced this work in the ITF and in RIPE, people said, don't bring human rights, you're politicizing things. But at the same time, economic values played in all these measures all the time. And even economic interests are part of BGP4. So financial incentives are seen as not political. I think this is structurally wrong and I think we should find a way in which we can develop a new social technical imaginary and a network ideology that makes network function in different ways. Especially in the light of the rising inequality in the world and the climate crisis that is right ahead of us. Please. So you also said that the philosophy and our norms of the internet development came as a reaction to how the telcos were structured. So where do you see the changes coming as a reaction to how the internet functions right now? Where do you see this coming from? And where could people kind of go to enjoy it, except for your lab? But where else do you see this coming to replace the old internet with something new? Well, interestingly, there is a lot of interest, actually, in telco development. So a lot of the new research we've been doing in the past years is actually on the roll out of 5G. And telcos have a benefit over the internet is that every 10 years, they have a new generation of technologies. So a friend of mine asked me, why was God able to create mankind in six days because there was no pre-installed user base? And where the internet is continuously backward compatible, telco networks can start over again and again and again every 10 years. So 5G is a huge push for a new networking, but you also see a huge discontent among people against the roll out of the technologies that's taking form of the burning of masts. So we see that there is a discontent and we see that in regulation, we see that in the actions such as burning of masts about how the internet political economy is functioning. So there is now a moment, a political moment, to rethink these structures. And I think we shouldn't let that depend on individuals because that's almost as saying fixing climate change happened with taking a plane or not eating meat. No, this is an industrial level problem that we should address also through native states and large movements. The internet was built through investments of native states where Gaia X went wrong was by giving it over to companies that took away the vision from developing actually a new network and standardized what they were already doing. So I think we should be courageous and find new ways, but new ways are never built in old structures. So if you're asking where does the new come from, I cannot say. But the new internet will not come from the internet. Whoa. Whoa, thank you very much, Nils De Nouveau for this fascinating and really intriguing speech. There was a lot of things to discuss afterwards. I hope people can find a way to you to do that. Yes? Thank you. Thank you very much. Give it a warm applause. And have a great, great MCH 2022 all. Thank you.