 OK, hello. Welcome. It's really a pleasure for me to introduce the presentation of Yuke Ferebeuf. Sorry, maybe for the bad French pronunciation. But he will present why we need to redesign internet to fight climate change. I'm very happy to introduce this presentation Yuke Ferebeuf is not only one of the experts of the D4S, the Digitalization for Sustainability. But he is also one of the leading people in the French think tank, the SHIFT project. So you published, I think it was four years ago or something like this, a very well appreciated report which was also going to a lot of media and press and it was bringing a lot of intention to, in particular, to the online video topic and was bringing the discussion also to Netflix use and this kind of stuff. So this was coming from this SHIFT project and he was the person behind. So I was well aware of your name since quite a long time, but I had never the chance to meet you. So now I hand over to you and at the end, you plan some time for questions at the end. Yeah. OK. OK. Good afternoon. Can you hear me all right? Yes, that's it. It will make things easier. So, yeah, why we need to release that internet to fight climate change? It's already used this title in some other presentations, in some other contexts and generally people are puzzled and why are you asking this question? Why should you ask it in this way? I think after three days of a bits and boi conference, it doesn't sound so peculiar. And we'll see especially why we have at the moment actually some systemic causes behind the fact that the digital environment footprint is not going in the right direction. And one of these systemic causes is not only the big tech, but the big tech business models. And so that's the purpose of this presentation is to explain why it is the case. I'd like to say that part of these results, you will be able to find them also in the D4S report that we presented on Friday and Saturday. And because there is a chapter on the business models. And also, there should be a specific paper published on that in a few weeks with all the details behind. So as it was said, what I'm going to do is to make a presentation that should last for about 35, 40 minutes. And then we have some time for questions and answers. However, please, in the course of the presentation, do not hesitate to ask questions for clarification. Not for discussion, but if I've said something which is not clear, please say it. That will be more efficient. So let's start. Yeah, as it was said, the project and actually myself, we did work on a number of reports. I'm not going to comment them all. And the first one is on the left. And that's the one that raised some attention from the media. I must say also that's probably one of the reasons why it raised some attention from the media is because it was published just one week after the IPCC 1.5 degrees report. So we were leveraging, actually, an existing context that was quite favorable. As a last report, we published a new project that was last year. And that was both an update of the prospective trajectories we built out in 2018 and also some thinking about 5G governance or more generally about new infrastructure governance. I'd like to mention that, I would say, one of the things we are quite proud of in the project is that probably we contributed to the fact that digital sustainability has become a topic not only for environmentalists in France, but it has become a law since it has been a law passed in 2021 called the Law for the Reduction of the Digital Environmental Footprints. And we did try the law, of course. So it was a long process, but part of these reports were contributors. And also two other reports that actually I did myself, one for the French Senate and one for the High Council for the Climate. And maybe even more importantly, digital sustainability has now become a regular topic within the association of the top French companies IT managers. That's the one in the digital sobriety. So and that has triggered the fact that now a number of companies are trying to understand how to still manage the digital transformation, but why trying to limit the environmental impacts of this transformation and also trying to make sure that this digital transformation is going to serve their energy or environmental transformation. So that's the good thing. However, I would say it's not because we published reports that the world is changing. And when we updated the figures in 2021, we had just to confirm that what we said in 2018 was still quite valid, actually. And the reasons behind it is that there are some systemic causes behind this trend. And to be complete, maybe you have noticed that the title of the D4S report we published is digital reset. That means that after two years of dialogue, we also would like to mention that I have a D4S colleague, two D4S colleagues, actually, a journal, please. And Mattias, thank you. We came collectively to the conclusion that trying to make progress through marginal changes will not be enough and that we need a fundamental reconfiguration of how digital systems are working at the moment. That's why we call that digital reset. So what I'm going to say, just two slides on climate change mitigation and not want to make a speech about climate change mitigation just to remember a few figures because I will use them later on in the presentation. Explained rapidly because it was, I think, largely covered in these BNB sessions why digital is part of the problem. And I will try to spend more time on why the big tech business models is a root cause of unsustainability. So that means there are two things behind. Root cause is not only the big tech, but also the business models. That means if these were not the big techs, but all the companies with the same business models, we would have the same consequences. And eventually, what is the way forward? How can we make sure we get out of this situation and what we should do? So I want to present these slides because anything I'm going to say later on, I'm not going to say it in absolute terms. I'm going to say it in view of the constraints we are going to face in the next 10 to 15 years. Because if we do not manage properly the 10-year period which is in front of us, well, the rest will become another story. And carbon neutrality in 2050 is nice. But the real objective for me is being able to halve the emissions by 2030. And that's the consensus today given the fact that decarbonation techniques or carbon capture techniques are far from being industrialized. We have to rely mainly on the reduction of gross emissions. And the benchmark of that is we want to not to go too much above 1.5 degrees. We should cut by half the emissions by 2040. And that should be valid for any industry sector, including digital. Second thing which is very important and to my surprise which is generally much less discussed is the fact that in order to be able to achieve this mid-term target halving the CO2 emissions by 2030, of course we have to substitute electricity generated through fossil energy with electricity generated via renewable energy. Good. Problem is, and we see that very much at the moment, is that the volume of electricity that we are going to be able to generate from renewable energy until 2030 is limited. It's limited because of our industrial capacity. It's limited because of the length time it takes to install electricity plants out of renewable energy in Germany, but in France, installing anything which is going to take things out of people and could be a nice view or something like that. You need 10 years actually to go through the process. So my point, and it's not he who's saying that, the International Energy Agency has been saying it for three years. If we want to be able to have CO2 emissions by 2030, we have to reduce globally the energy consumption by 2030. And on average, we need to reduce it by 7% between 2020 and 2030. And the effort needs to be higher in developed countries. For instance, in Europe, it is 15%. So why am I stressing on this point? Because any carbon neutral strategy that does not also include a strategy on how to reduce the energy consumption is not going to work. It might work maybe on a local basis, but if it works, it would make things more difficult on a global basis. So that means that also on the digital side, the digital energy consumption at least should not be increasing. On the slide? OK. So the energy needs to reduce, but this applies. Sorry, this is showing an increase in electricity generation, but the total energy supply, like heat energy, stuff like that, that reduces, but the electricity use would be going higher, yeah? Yeah, yeah. Yeah, electricity is going to increase, but the total energy supply must decrease. And that generally is the missing part of the speech. OK, so raw figures, digital footprint today between 3% and 4%. So higher than airline industry. But maybe more importantly, GHG emissions are growing at a rate of about 6% a year. And so 6% a year is about exactly the contrary of what we should be targeting. We should be targeting for minus 6% a year if we want to divide emissions by 2. And it's increasing by 6%. So that's why I said at the beginning, the trajectory is not on the right path. If we look at the footprint, clearly the footprint due to end user devices is a larger spot, about 75%. It depends on the countries, actually, but on a worldwide basis. And within this part, the production part is very important. The carbon which is embodied in the equipment during the production phase is generally higher than what is going to be generated during the use phase for an end user device. For a smartphone, it's about 85%. 85% of the carbon footprint of the smartphone, if I keep it for three years, is the emissions which were generated when it was manufactured. And that's, again, a worldwide average, because when you manage to reduce the carbon intensity of the electricity, the carbon emissions of the use phase are going to decrease. You'll keep the 7G. And so therefore, the proportion of the production phase is going to increase even more. And I didn't show it there, but if you looked at any of these components, you would say that all of them, actually, their contribution to GLG emissions or to energy consumption is increasing. There is no one component which is especially responsible, I would say, for the growth of the digital footprint. So that's one thing. Second thing is, of course, when you produce more, and if you have not in place a circular economy, well, you get more materials at the end of life. So in 2019, there were about 55 million tons of electronic devices that were, let's say, going somewhere. And I'm saying that like that because we actually know where this somewhere is, only for 15% of these 55 million. The rest of them is a gray area. But a good part of this gray area is something like that, which is in Ghana, and where, ultimately, these electronic devices end up and where people will try to get out of them where they can. And, of course, with a lot of prejudice both to the health of the local populations and everything. So not a very nice feature. And so if you attended the different report presentation, but Lawrence Hilty, who is one of our colleagues, explained that if we are in these situations, not at all. Because there is not enough progress in terms of energy efficiency. No, probably the digital sector is the one which has been making the most progress in terms of energy efficiency in the past 15 years. And you can see that, for instance, when you see actually how many computations you can do per kilowatt for the different generations of processes. You can see that each generation is almost bringing a progress of one of the magnitude compared to the former one. So if we have this growth of environmental footprint, not because of energy efficiency, it's just because we have too many digital volumes. And you can see that on this graph. Actually, the CO2 footprint is going to result from the product of the energy consumption coming from digital and the carbon intensity of this energy. Then the energy consumption is going to result from how many people are using digital. And this is increasing but slowly now in the past few years. But more importantly, it's going to result from what is called here energy intensity of technology, which means in practical terms how many end user devices you own per person, how much data traffic you are generating, how long you are keeping your existing devices, and so on, and so on. So this is what I said relates to technology influence. And this is actually the factor which has been increasing the most. And I must say, I don't have time for that, but I can answer questions in the end. I must say where it has increased the most is also where it is already the largest. In other terms, in the US a few years ago, there were about eight digital devices per inhabitant. At the moment, there are about 13. So 55% increase. On a worldwide basis, probably there has been an increase, but only of 15% or 20%. So where there is already a very high usage of digital is also the places where the digital technology influence is squeezing the fastest. And that's some examples. So here on the left, you have what is called here the stock or the installed base of digital devices. So we have about, at the moment, about 20 billion digital devices installed in use in the world. And if we believe the analysts, sector analysts, we should have close to 50 billion devices in 2030. So more than doubling in 10 years. And another example is the growth rate of the traffic in the mobile networks, which is on a worldwide basis on the past three years in the range of 45%. 45% a year. Just huge. And as you can see, again, according to the analysts, there is no thing which says that this trend should stop in the near term. And of course, what is true for the installed base is true also for the sales of new equipment. And at the moment, we are selling about 4 billion digital devices a year, and probably be selling twice as many in 10 years, which means that between now and 2030, we will sell about 70 billion digital devices. And if we have 7 billion people on the planet that means 10 per person. And what is happening in the digital world is that at a given time, you have a number of different equipments that are in use. And then some of these equipments become less popular. But actually before they become less popular, you have new types of equipment which arrive and which are actually a new growth driver for the vendors of digital devices. And that's what is happening at the moment with the internet of things where we actually were going to equip virtually any non-digital thing with a digital module to make sure that it can be connected to the internet. So there has been a lot of discussion, but why do we have that? It's because people want new services. It's because the vendors are still in the linear economy, and so they need each year to sell more devices than before, and so on, and so on. Actually, it's a bit more complex than that. There is a lot of interactions now, and digital is already so much present in our daily lives that actually the dynamics behind this trend that I've described are very complicated. So it has really become what can be called a systemic issue. And some characteristics of this systemic issue. The first one is very important. That probably even you may be probably a few things that I said you didn't know them. For instance, who knew that a smartphone like this, this cardboard footprint is about 80 kilos? Who knew that? OK, two people out of 30. And I think you are probably more informed than the average citizen on these type of things. So there is a complete unawareness of the sustainability aspects of our digital things and usages at the moment. So there is nothing, maybe it's a bit trivial example, but when you have a car and you decide to go in Germany, you can do more, but you decide to do 140 on the Autobahn, rather than 110. Then when you have to go to the gas station, you will see that you need to put 15% more gas into your tank. So it's 15% more expensive. So after a while, maybe you decide, OK, maybe I will not go down to 110, but I will go down to 120. So what I mean is there is a feedback loop that helps you to regulate your car usage. At the moment, there is no feedback look at all in digital, especially with the fact that a lot of services now are either offered for free or seem to be offered for free. Or they are offered for a fix on, and then you can consume as much as you want. So there is no mechanism in place that is there to facilitate this self-regulation. On the enterprise side, and I feel a bit guilty about that because I was one of them not so long ago, the companies have been briefed by digital vendors on the fact that to achieve future growth, they need to transform digitally, and they need to do it faster than the competitor because if they are late, they will lose market share. And we've been quite effective actually because now every company almost is working on its digital transformation, and they have associated so much this digital transformation effort with, let's say, future profits or even the future of the company. If we don't do it, we are going to lag behind and maybe we won't exist in five years' time. That it's not natural at all for them to think or to admit that this digital transformation could be detrimental on the environmental side. And then the third thing, a bit the same phenomenon. Policymakers are very worried about anything that would prevent digital to grow because they think that it could affect negatively the GDP of the country and the growth of the GDP. Okay, but there is one other root cause which I like to comment in more detail today which is the fact that the big tech models that rely on audience maximization is actually intrinsically incompatible with sustainability. So what makes it worse is that these big tech are completely dominant. And that's 2020 figures. 2021 figures would be even more threatening. But the total market capitalization was in the range of $10,000 billion. And as you can see, these big tech are virtually the top ten companies in the world by market capitalization. So anything they do is going to have a major impact and anything they want is also going to have a major impact because they have a lot of power to achieve it. And these dominance actually, we can find it also almost on the physical side of things. In 2021, six companies have generated 57% of the total internet traffic in the world. 57%. Six companies. And in some countries it's even worse. I don't know in Germany, but in France it's probably around 70%. So, and that's the market share. But if you look at what has been going on in the past three, four years, you will see that actually these six companies are responsible this time for 75% of the growth. So virtually, not only they drive the market, but they create the market. And you can see the same thing on the smart smoker market where they have actually 70%, 80% of the market share. And their dominance creates a lot of problems and competition, privacy, social toll consequences and so on and so on. It's not my point today, but it creates also a lot of problems on the environmental side and why. Because actually there are probably one of the few companies in the world whose revenue energy intensity is growing. So what is the revenue energy intensity? It is actually the energy a company needs to generate one euro of revenue. And what is happening is that each year, for instance, Google is going to use 4% more energy to generate this euro of revenue. And why am I saying that it's probably one of the very rare companies that are in this situation? In the past 10, 15 years, the revenue energy intensity at world level, which means the energy intensity of the worldwide GDP, has been decreasing at about 1.5% a year. And it's actually something which is necessary if we want to be able to reconcile as much as possible economic growth and reduction on environmental footprint. It's called decoupling. So here, it's not at all decoupling. It's a coupling in the other way. Each year it's getting worse, actually. And why is that? How, for instance, does Google make most part of its revenue? Well, what they do is they have a lot of users which go on their platform, Google, YouTube and so on. And when you go there, well, maybe you can provide some personal data but you also provide, without knowing it, a lot of behavioral data. And these data, Google are going to use them in order to build some information that then they will be able to sell to third parties that will use this information to try to sell services and products to you. That's their business model. And so, what is the... What's going to make them successful? What is going to make Google successful? It's its ability to attract as many users as possible on its platform and to make these users stay as long as possible on its platform. And how can Google achieve that? Well, it can achieve that by making its platform always six-year, always more attractive than other platforms. And for that, what are they going to use? They are going to use data. They are going to use videos, especially at the moment, probably virtual reality in a few years. So they are going to use actually some data as materials to build an environment that is going to make you stay as long as possible and to come back as often as possible. And each year, they need to enhance these environments to make sure that it still remains attractive and more attractive than alternatives. And when they do that, they are actually responsible for a huge augmentation of traffic. You see, 42% a year for Google, 60% a year for MetaFacebook. It's gigantic growth rates. So the consequence of these business models is that not only is going to increase their own direct energy consumption, the direct energy consumption what it is, it is the data center's energy consumption. Google's energy consumption is increasing by 25% a year. 25% a year. So it's increasing a bit more rapidly than its revenue. And they have that even though their data centers are probably the most efficient in the world. So again, it's not a problem of energy efficiency, it's a problem of influence. Now, since especially two years, a lot of these big tech have announced to the market, no problem, we have plans to become carbon neutral in the fall of 2040, 2040. But don't worry, we are going to cut our emissions. And actually, when they do that, they are only speaking about what is called SCOP-1 and SCOP-2 emissions. Are you familiar with this? Who is not familiar? So when you try to establish the carbon footprint, for instance, of a company, you have several, let's say, SCOPs that you need to investigate. The first one is how much fossil energy are you consuming directly? Means with your cars, with your trucks, with your central heating, if you use gas or petrol and so on. SCOP-2 is basically how much electricity do you use? Because we know that electricity is still... 60% of electricity in the world is still produced out of fossil energy. So consuming electricity at the moment creates emissions. So that's SCOP-2. And in many countries, companies are obliged to measure and to report only SCOP-1 and SCOP-2. So what is SCOP-3? SCOP-3 is actually, in the digital world, especially two important things. It is the carbon footprint of the equipment you are buying, because you have not produced them yourself. So it's not in your SCOP-1 and SCOP-2, but you are buying it. And there is, as I said, a lot of embodied carbon in these equipment. And the second important sub-scope, if you want, is the emissions that are going to be emitted by the services or the products you sell when somebody is going to use them. Because in SCOP-3, we consider that the company should be responsible for the emissions that are going to be due to the usage of the services and products they sell. And in digital, it's very important because SCOP-3 is generally 10 times as big as SCOP-1 and SCOP-2. So when Google and the others say, we are going to divide emissions, we are going to be carbon neutral, and if you remember, that means we have to divide emissions by half by 2030. They are talking about SCOP-1 and SCOP-2. So time is running, so I'm not going to do all the details, but I guess you will have the presentation. If you do the math, you will come to the conclusion that indeed, even though they will keep having electricity consumption grow at a pace of 15% to 20% a year, if they do equip, for instance, all the sites with renewable plants producing their own electricity, which is what Google is doing at the moment, they will indeed manage to reduce their emissions. But in the meantime, their electricity consumption will have increased enormously, you see. At the same rate, it will go from 70 terawatt to almost 350. Okay, so multiplication by 5. But it will be also true for the rest of the valued chain. The rest of the valued chain is not only the energy that's going to be consumed by you and me when we are going to use Google services, but also the energy consumed by the networks that are necessary to establish the connection between the user and the Google data center, for instance. So the rest of the valued chain, of course, the energy consumption is much bigger. It will increase a bit less, but it will still increase almost by a factor of 4. And actually, the results are right. The scope 1 and scope 2 will be reduced by 50% if they do that, even though they will have multiplied the energy consumption by 5. Okay? But in the meantime, the carbon footprint of their valued chain, that means their scope 1, scope 2 and scope 3 will have increased by 50%. Okay? That means that there are two comments. One is for the reason that is described there, there are business models and the growth profile is incompatible with sustainability constraints. The two, the, let's say, sustainability plans, they have announced, makers believe that they are not anymore afraid. So they can use that to legitimate the fact that they can keep growing as they have been growing. So it's a double threat because actually, yes, they will be able to show they were right, but in the meantime, they will have continued gravitating actually the environmental footprint of the digital sector. So as we have only 5 minutes left, so what can we do? I think as I said during the DeFloas presentation, the first thing we have to do is to make big tech smaller because if we don't do that, of course we can push and we can support alternative initiatives and alternative platforms. But I went to several workshops on these in the past two days and what I heard is that it's very difficult at the moment because they are fighting against giants and also it's very difficult both to receive enough financing and also enough, I would say, publicity to help them become, just to become a viable alternative. Because if you don't know they exist, you are not going to use them. So the first thing we need to do is to make sure that the big tech themselves become smaller and that's competition law and so on and so on and maybe the continuation of the DMA, DSA regulations in Brussels. But it's also making sure that their business models as such are going to be made less attractive and that's about data regulations and also about sector regulation and for instance we should avoid these dual-sided markets where actually some players manage to externalize large parts both of their costs and both of the financial and environmental costs but to the detriment actually of the rest of the society and of the economy. So probably it's a very limited list but one way which seems quite interesting to me is the idea of cooperative platforms, especially in Europe. So we can think about let's say cooperative startups that from the start adopt a pure digital model but with a cooperative legal arrangement. But we can also see the 220,000 cooperatives that exist in Europe and see how they can transform digitally and become themselves platform but remain cooperatives. It shouldn't be so difficult actually. All the technology is there. These cooperatives at least locally are well-known players. They have customers and so on. So just have to make sure they are being provided with the tools to become more efficient and maybe to upscale the operations. So that's the alternative we should be pushing for. As I said, the public policies which are needed first is to make big tech smaller but also the hell is in the details on that. For instance, when you first go on Facebook let's say you're being asked whether you agree on the terms and conditions for the use of the service. Of course 90% of the people are not going to read the terms and conditions and if you do that you will have very long pages of legal terms and after a while they don't care. I'm going to accept. But actually the first thing is what do they do? They establish a contract between you and them and it gives them the ability and the right to use the data you provide for free in the way they want basically. So it's very difficult to check whether it is what is exactly in place so I think we have to enhance the existing data regulation and forbid the acquisition of personal data by default. That means there should be an explicit authorization on a detailed set of data that should be given and also last but not least we have to make sure that the technical architecture of the web is going to evolve according to directions that will make all those things easier and especially giving the ability to each individual user to have, I would say physically, the full control of the data. Sorry I would have been a bit long if you still have some questions. I'd be happy to answer them. Yes, I guess we have time for maybe two, three questions. Hi there, Chris Adams, Green Web Foundation. I have a question. Are there any other sectors we can learn lessons from to address some of the problems you described about a small number of very, very powerful players or people having something they had for free that they might not want to pay for? Because there are other sectors we've had to deal with the same kind of consolidation of power and influence previously. I'm not sure I have the answer like that. I think one difficulty we have with digital is that it's a sector where technology and usages are changing very fast and it makes a hard time for policymakers because they're always behind. They're still trying to fight the battle before the last battle but the new war is coming and they're not seeing it yet. And I think we're going to see that with the Metaverse thing which is full of problems. So I'm not sure we can really derive any lessons from another sector. Another question? It's too late. One question. So on the last slide you mentioned Web 3 as one of the options. I find that very interesting because Web 3 I would connect to the blockchain industry decentralization in that direction which I don't think is less energy-consuming than what we already have and I think there are alternatives that actually exist. There's something called Swall Web which is exactly the counterpart to Web 3 which then goes like degrowth in that sense. So maybe you want to comment on the Web 3. I used Web 3 in a generic term because there are still a lot of different approaches of Web 3. And we should certainly avoid Web 3 to create new monopolies in the place of other monopolies certainly. No, I was saying that as there is a working process on how to evolve the internet architecture these principles should be at the core of the redefinition of the architecture. Now is blockchain a solution or not and which type of blockchain and how can blockchain and help that one of the many technical questions should be resolved but it should not be seen as a religious option. Thank you. Another question, maybe one last. That's not the case. So once again big applause and thank you very much for your presentation.