Vinton Cerf:
"Internet XXIst century: the tidal wave"

The father of the Internet provides us the keys to understand the Net, and presents the new challenges to be solved through its evolution.

Well-known as the father of Internet, yet he always deny to be so, Vincent Cerf is the author of protocol TCP/IP, the knowledge of which provided Internet’s success and its revolutionary effects. Fierce defender and activist of the availability of new technologies in Internet network, not only including countries, continents and individuals but with a special awareness for disabled people.

Invited by INISI (Institut de Dret i Societat de la Informació), Vinton Cerf visited the UOC on May 23rd and he gave the following talk on the framework of the IN3 (Internet Interdisciplinary Institute):

Internet XXIst century: the tidal wave

Let me begin by describing for you probably one of the most pressing problems that we face in the Internet today: hackers. There have always been the vandals, hunts and now we shouldn’t be too surprised that hacking is still a problem. Some of you know that some days ago in Tarragona we unveiled a new street called ‘Internet street’. In our discussion with the major of Tarragona, we decided that we need another street called ‘Hacker Alley'.

It would be helpful to understand the power of the Information Revolution by going back to understand the power of the Industrial Revolution and how much it changed the way we live and work. I am going to draw a kind of analogy: when the Industrial Revolution first started, the way in which we produced mechanical work was to find a river and build a water wheel on the river and then use the mechanical work of the water wheel to drive belts and polleys to make engines function so we could manufacture things faster. But that limited where we could place our manufacturing facilities because they had to be next to the river

So the next invention was the steam engine and allowed us to move the manufacturing plants to almost any where we wanted to put them as long as we could build a steam plant so we needed water for that but we didn’t need running water.

Then the next invention was electrical power generation and electrical motors. This is a major transformation because now wherever we could send electrical power we could put an electric motor to do mechanical work. Moreover this motor could be as large as or as small as we wanted to drive elevators up and down or to work in electric clocks. In fact, today each one of you has many hundreds of motors that are working for you 24 hours a day and you don’t have to pay any attention to that. You have motors in your kitchen appliances, you have motors in your cars, and also in your refrigerator there’s a little motor that drives the compressor to keep your food cold.

Just like the electric motors, the computers and the software in them and the information is there to multiply our brainpower. And because of that comparison I think that the information revolution is going to be just as powerful maybe more powerful that the Industrial Revolution that preceded it.

Most of us don’t lie awake at three o’clock in the morning worrying that the electricity is going to go out because we assume that the electrical power system is going to operate most of the time, we assume that most infrastructure, we assume the telephone system will work, and we get very upset when these things don’t work, when the power goes out, when the telephone system doesn’t work and when the routes become congested, we are very annoyed because we have come to depend on the infrastructure. Internet is something we’re beginning to depend more and more. The analogy I want to draw between the Internet and the electrical power system is that instead of electrons going through wire the Internet moves information through the network and it delivers the information to computers that are running software, so these computers are like the electrical motor. There’s software running in those computers 24 hours a day, so they work for us while we are doing something else. Just like the electric motors, the computers and the software in them and the information is there to multiply our brainpower. And because of that comparison I think that the information revolution is going to be just as powerful maybe more powerful that the Industrial Revolution that preceded it.

We use the term ‘Internet’, we say ‘The Internet’ as if it was ‘The network’ but it is really hundreds of thousands of networks that are connected all around the world and they work because they all use the same communication protocols.

The first two protocols that were designed were called TCP, Transmission Control Protocol and IP, Internet Protocol. The second is very simple and it can run on any underlying transmission system. At first, we did not know how telecommunications technologies would come along in the future, we did not about things like RDSI, but we knew we wanted to design something that could operate as simply as possible and could be the basis for all. All that Internet architecture requires is the underlying communication system that takes a bag of bits and moves them from point A to point B, with some probability bigger than zero. As a result, that very simple requirement has allowed us to put this Internet protocol has allowed us in literally every communication system that has been invented in the last 25 years.

Most of you know in detail how the Internet work but for those who does not, it is important to have a clear understanding of how the Internet functions if you are trying to make laws about the Internet. Otherwise, you may be tempted to make laws that cannot be in force.

Let me start by reminding you how ‘addressing’ is in the Internet. This system works just as the telephone numbers except that in the Internet you need 32 bits addresses, that is enough address space to label up to 4 billion devices on the Internet. The important point is that we have had to evolve the structure because we have an increasing number of networks to interconnect.

In 1973, we didn’t think that we needed more than 128 networks, because at that time that was more than the number of countries and we didn’t think that we needed more than one network per country, and besides this was simply a research program for the US Defense Department. We could not afford to build many networks because they were expensive. Now we know we need many more than 128 networks. Today we have enough room for 4 billions.

We could draw an analogy between Internet packets and postcards because everything about a postcard applies to an Internet packet: there are receiver and sender addresses, these addresses are formed by numbers ranging from 0 to 155 and they represent bits of information. There is also the version of the Internet protocol is carrying and a content.

So, let’s think about an Internet packet as if it was a postcard. What happens to a postcard when you put it into the postal system there’s no guarantee that it would come out, and that is also true for an Internet packet. This is what is called a ‘best effort system’: if you put two cards into the postal system to be delivered to the same destination they don’t necessarily come out in the same order you put them in. This is also true of the Internet packet. Internet does something that the postal service doesn’t do; sometimes it would make a copy of the Internet packet and deliver the two of them.

... the networks as bucket brigades, when people would have buckets full of water to put out a fire. The Internet system has these computers called ‘routers’ that hand buckets of Internet packets to each other, one by one, until they get from the beginning to the end.

We can think of the networks as bucket brigades, when people would have buckets full of water to put out a fire. The Internet system has these computers called ‘routers’ that hand buckets of Internet packets to each other, one by one, until they get from the beginning to the end. So, there are really hundreds of thousands of bucket brigades in the Internet, each brigade being a network, and they are all connected together to form the Internet.

You would be quite justified in saying that there is no obvious reason why a network like the one I just described would be very useful: packets are not guaranteed to come out, they don’t stay in any particular order, sometimes they get duplicated. So, how can you make something useful out of that? The answer is that you have to put on top of that very simple Internet protocol a procedure to deal with the problem of losing and duplicating packets. This notion of layering is pretty important to the Internet architecture. There are many layers and procedures in the system and each one of those layers is typically another protocol.

The second most important protocol is the TCP, it stands for Transmission Control Protocol, and it is in charge of fixing all the weaknesses of the underlying postcard- like system.

How does the TCP wok? Let’s imagine that I’m going to send a book to Manel through the Internet, suppose I was going to send it through the post office the only way I could send it is through postcards.

Not every postcard can have a page or a page number. Some pages take more than one postcard and as I cut the pages up, they are not going to stay in order. So, how is he going to put them back together? I’d better put a number a sequence number on every single postcard so that Manel can put them back in the right order. Then we have to remember that some postcards are going to get lost so I’m going to keep copies in case I have to send new ones. How do I know if I should send the copy? I would ask Manel to send me a postcard saying, for instance, ‘I got all the postcards up to number 326.’ Then I remember that the postcard he’s supposed to send me may get lost. So, in the end, I decide I have to have a kind of timeout that says if I’m hearing nothing from Manel at all I had better start sending copies of things that have not been acknowledged. That is about all there is to this TCP protocol except it works in Internet packets system instead of postcards and it works at a hundred million times faster than the post office.

Suppose we started with the book that had 200 pages and we cut it into 2000 postcards. We take all these postcards to the post office and by some miracle all 2000 postcards get on the same day in Manel’s mailbox. There was not enough room for them, some of them would fall, disappeared, and have to be retransmitted. So, I might have an arrangement with Manel saying I won’t send more than 100 postcards at a time, that way I won’t overflow your mailbox. That’s how TCP/IP works.

That is why we have a whole elaborate system, the “domain name system” that translates the domains into addresses, that in fact are numbers like the telephone directory, you look up a person’s name and the directory tells you the telephone number to call that person.

Other details, for example, in the network we have labels that have the different computers on the network and we give them names domain names (.edu), (.com), what’s important is that domains are for human consumption but the network doesn’t know what they mean nor where these domains are, they do not know where the www ‘whatever’ is, they need to look that up in the directory and be able to read it. Each domain must be translated into numbers: 208, 234, etc before the Internet knows how to send the packet through the net. That is why we have a whole elaborate system, the “domain name system” that translates the domains into addresses, that in fact are numbers like the telephone directory, you look up a person’s name and the directory tells you the telephone number to call that person.

... more and more trademarks are domain names, and, worse, some domain names are being registered as trademarks.

There’s another problem in trademarks and domains because now that the Internet has become a commercial activity, more and more trademarks are domain names, and, worse, some domain names are being registered as trademarks. Trademark law does not require that a trademark be unique, that is own by only one organization. There can be many companies that own the same trademark. In the US there are two companies that have the same trademark MCI, one is my company, a telecommunications company, and the other is in another state that makes buses. It’s all right under the trademark law for the two companies to have the same name, because these two things - telecommunications and buses - are not confused. The trademark law allows ambiguity but the Internet domain system does not allow ambiguity: every Internet domain must be unique. It’s like 800 numbers, that is, free phone numbers; they have to be unique.

We are in the middle of the Internet gold rush. People are rushing around looking for gold but we’ve learnt some lessons from the past: the people who make money in the gold rush are not necessarily the people who were looking for the gold but the people selling tools to the people looking for the gold.

Some people that are looking for gold in the net are like the people looking for gold in the mines. There is a very peculiar business model which has a rapid evolution, you started a company, and run the company for about one year and open up the company to public ownership by having an initial public offering of stock and you go and buy something that is worth something so that you can justify the value of the stock.

... a company cannot work losing money for many years.

We are in the middle of the Internet gold rush. People are rushing around looking for gold but we’ve learnt some lessons from the past: the people who make money in the gold rush are not necessarily the people who were looking for the gold but the people selling tools to the people looking for the gold. That’s how MCI made money, selling tools and it’s a very good business, we don’t have any complain about this. Some people that are looking for gold in the net are like the people looking for gold in the mines. There is a very peculiar business model which has a rapid evolution, you started a company, and run the company for about one year and open up the company to public ownership by having an initial public offering of stock and you go and buy something that is worth something so that you can justify the value of the stock. That’s an interesting business model but we’re now finding that the analysts in Wall Street are asking very good questions about Internet like it would be a very good idea if these companies made more money that the money they invested because a company cannot work losing money for many years. These analyses are really good: even an engineer can understand this kind of business.

Internet as a commercial enterprise is still pretty young, it was not until 1990 that people were allowed to do anything commercial in the Internet and there is still a lot of people experimenting. Some of the business created that don’t seem to make much sense but we have to think that they are experiments. We will have to wait to see what kind of business will work on the Internet.

It would be useful to spend a little time on how big is the network and how fast it is growing. Just to give you some data, for the last two and a half years, from the middle of 1997 to today. In the middle of 1997, there were about 1.3 domain of level 2, registered in (.com), now there are more than 10 million; two and a half years ago there were 22.5 computers on the net, not laptops or PCs but routers, web servers, email servers and so on whereas now there are 72 million and for 2001 January I would estimate 100 million. The number of countries that had access to the Internet was about 109 in 1997 and now is 118.

... many companies want to appear to be global and do not want to be associated with any particular country so (.com) has become a very popular place to register.

The network is not uniform but more available in some countries than in others. The number of users in the net was about 50 million two and a half years ago and is now 276, or 300 million in May. To put this in perspective, there are many terminations on the telephone system, so Internet is still very small compared to the telephone network, but the telephone network tends to grow at 5 to 10% per year, not counting mobiles, that grows at 30% or 40%, Internet has been growing at 80% to 100% per year since 1998 and it continues to grow at the same pace. An interesting statistic that I only learnt yesterday: in (doc.es) there are 18,000 domain name registrations and in (.com) there are 200,000 of Spanish origin. These statistics are very similar to what we have in the US, where there are a few registrations in (.us) and great many registrations in (.com). What this suggests to me in part is that many companies want to appear to be global and do not want to be associated with any particular country so (.com) has become a very popular place to register. There are some local restrictions in (.es) that make it harder to make a registration in domain names, maybe we can come back to that during our discussion time.

In terms of the locations where Internet is in use, the statistics I have say than less than half of the Internet users are in North America, 5 years ago it was completely different, and most of the users would have been in the US and Canada, but now 72 million of the 275 users are in Europe and this is interesting because the European population of Internet users was about 54 million some months ago. In Asia, in the Pacific Rim, 55 million users. The absolute numbers are disappointing because the population in the Pacific Rim is over 2,000 billion in absolute numbers in Japan, India, China, Indonesia, Malaysia some of the most populous countries, so 56 million out of 2,000 is a pretty small number but there are a number of reasons for that not to list that the telecommunications infrastructure in some of these countries is still very weak, for example, in India or in China.

Latin America is starting to wake up, six months ago there were about 6 million users now there are almost 8.8 million. In Africa, the population on the Internet has doubled in the last six months but still 2.5 million is a very small number, there are almost a billion people in Africa so clearly 2.5 million is small. Most of these 2.5 million were in South Africa, Egypt, Morocco or Tunisia, the rest of them are in small numbers in other parts of Africa. Some of these countries have a very reduced number first, because of the very weak telecommunication infrastructures and second because of the lack of money available, a situation that doesn’t help to sustain Internet.

In the Middle East most of the users are in Israel but the surrounding Arab states are beginning to pick up and use the Internet as well.

Just to show you that there are people that want to get into Internet no matter what. Now I am going to download something. It’s a 14-megabyte Power Point file and it may take a while. It’s a picture that I got from two people in Uganda, in Africa. These people had a cellular power converter and a satellite antenna and a personal computer. This little village didn’t have any power and it didn’t have any telephone service but they wanted to be on the net so they used the solar power converter and the satellite antenna and they took a digital picture of themselves on the roof adjusting the antenna and they email this to me, so a few weeks ago I got this picture attached to an email saying ‘We’re on the net!.’ It’s important to see that if you’re really determined to get on the net, you’ll find a way to do it.

... beyond 2010, half of the world’s population will be in the net (...) that would mean the Internet would be about the same size as the telephone system in about 6 years.

My estimate is that  beyond 2010, half of the world’s population will be in the net, we will get to test this theory and see whether I was right. Other estimates I have made in the past is that the number of devices on the net will be about 900,000 million by 2006. This includes not only desktops or laptops but also other kind of devices. That would mean the Internet would be about the same size as the telephone system in about 6 years. I have talked to Nokia and Ericsson and they’ve told me that they expected to have 1,500 million Internet enabled mobiles on the Internet by the same time in the year 2006. It started to scare me a little bit because the address space is 4,000 billion addresses there’s a lot of wasted space so by 2006 we will be under very great pressure to have a new address space which is bigger.

Let me now talk about some technology that is creating some problems at the policy levels. We are being asked today to carry video, telephony and other information on the Internet, sound, for example, it’s not new to carry sound, as far as in 1975 I was working on that but in those days, the speed of the backbone was 50,000 bits per second, that is, slower that any telephone line today, but then it was all the capacity we had. So, in order to carry sound on the net, voice on the net, we had to compress that voice signal and digitize and compress it and we got it compressed at 18,000 bits per second using some special techniques: there was only one little problem, anyone who spoke through that compression system come to sound like a drunk Norwegian. So, the day came when I had to demonstrate to some generals in the Pentagon and we asked one of the Norwegian researchers to be the speaker and it sounded just like him so we didn’t tell the generals everyone’s voice would sound that way.

In the last 25 years the net has improved a lot, we have more capacity on the net and we can transmit very good quality sound. So, today there are about 8,000 radio stations that transmit their sound through the net. This raises an interesting business problem: in the normal radio stations, radio audience are defined by the power of the transmitter, how far away it is received and the geographic definition for that radio. But on the Internet you can be anywhere in the world so now the definition of your market is not geographically defined anymore.

As for video, it is very similar to radio on the net, it takes about 400 kilobits per second to deliver a quality video. There is a problem there because this is 8 times faster than what you can get with a normal telephone line. So, in order to deliver good quality video through the Internet we need higher speed access lines than we can get with the telephone network , there are some new technologies available: digitable subscribable loops, cable modems and others that work at million of bits per second but we won’t have access to those alternatives unless the telephone companies and cable are able to make available that kind of facility.

We are seeing an increasing amount of these in the US, I don’t know how active the telephone companies are here regarding the implementation of these technologies, we can talk about this later. Anyway, is it possible to send video through the net? Some people try even at lower rates than 400 kilobits per second but the result is not very good. Every time I do that I expect to see a message congratulating me for having turned my laptop into a television set.

The problems we have are not related to the bandwidth, with the capacity, but- do you remember the brigades? - , with the nature of the net, the time when you send something and the time that it gets to the other side. If you ever talk to someone over a satellite loop, there is about half a second of delay, and on the phone also, on the Internet this half second of delay feels like a long time and I can’t help having the feeling that the person I am talking to does not agree with my idea and is trying to figure out a nice way to tell me. So, there’s a new social effect that the delay introduces caused by the voice on the Internet. Internet should be re-engineered so that we can solve this situation. Then, we will be able to transmit our voices the same way we transmit other kinds of information.

Traditional media like cable, satellite, radio, are capable of carrying Internet packets so Internet is in both directions. The two are not colliding but mingling together. That is causing great regulatory confusion ...

I just want to mention one other thing, the traditional media, radio, television, telephone are showing up on the Internet so Internet is carrying those media. Traditional media like cable, satellite, radio, are capable of carrying Internet packets so Internet is in both directions. The two are not colliding but mingling together. That is causing great regulatory confusion because, in the past, regulators could distinguish between television broadcaster, cable caster and radio, telephone company, but now that everything is carrying Internet in a uniform way is not clear whether you should regulate.

Remember the statistics that not counting the mobiles, there will be 2,500 million of devices in the Internet.

Two graduate students at the University of Massachusetts built a two-chip web server, one chip was used for TCP/IP protocols and the other chip delivered the files to the web server, the typical HTML. So, that whole web server was the size of two American quarters, a pretty small device of hardware, with a low cost. That means that any graduate student can design devices for the Internet.

Let’s talk now about Internet-enabled products. If your family is like my family, the communication sender is in the kitchen and it consists of paper and magnets we put the paper in the refrigerator door and hold it with magnets- the Internet-enabled refrigerator is an improvement because you can put email on the web. Another example is an Internet- enabled picture frame. When I first saw it at a consumer electronic show I thought it was as useful as an electric fork, but it turns out that this is quite clever. You don’t have to program it or worry about it. If you are a grandparent and you don’t want to learn about the Internet, but you want to see pictures from your grandchildren, you can get one of these frames and connect it to get pictures from your family automatically every morning.

There will be an increasing number of small simple devices. We also have these new computers, the Palm Pilots, which are in fact mobile telephones enabled to connect to the web and send and receive messages. The Japanese are now creating games with capacity for the Internet, so if you’re at home playing with friends who are separate from you, you cannot see each other, it would not take much to put a microphone on the videogame so that players can send and receive sound. A company did that and they discovered that the children that were able to talk to each other while they were playing the game were more effective than the ones who had to stop and type to each other.

My prediction is that the business world is going to adopt video conferencing with this videogame technology because the videogames are very inexpensive compared to the very elaborate videoconferences.

Going back to the example of the refrigerator, it could have a bar code scanner so that it knows what you put inside, you could go on the net to find recipes  and to cook with what you still have in the refrigerator, it may also know how long the item is been there, so you would get an email from your refrigerator saying you have to cook something in particular because it’s been there for 3 weeks.

When I was in Japan they showed me a bathroom scale that was Internet enabled. When you step into this scale, your doctor receives your weight but, what happens if the refrigerator gets the same data and it doesn’t open any more?

I was also very interested in an enabled car, it was a Mercedes Benz, they had two liquid crystal displays in the back seat for the children to play games or to get on the net. There was also a display for the driver and the car had a global position satellite receiver so that the computer knows where the car is. That means that you could have programs that you can reach out in the Internet and get useful information back from databases that are geographically indexed. So, you can ask questions like ‘where is the nearest restaurant’, ‘where is the nearest hospital’, ‘can I make a reservation’, ‘can you show me directions for getting to where I want to go’, and so on.

For many years I was comfortable doing my email while I was in the airplane because the display was so bad that even people sitting next to me could not read it. Now the displays are very good indeed. Now, how could I protect my privacy while I’m reading my email? I got the idea that we could have special glasses that project an image on to your retina. There are people experimenting with that in the MIT, but I got to wandering what it would be like to be sitting next to someone on a plane who’s using these reading glasses and pointing the 3D display. It will not be a very comfortable situation for anyone.

We don’t need wires any more to be on the net. Nowadays, there are all kinds of wireless devices. There is a great deal of development now not only radio but also infrared. From the GSM design  with only 9.6 kilobits per second, which is not bad for sending an e-mail but is not very comfortable for surfing the web, and the new chip PRS we will have a bandwidth much better, 100 kilobits per second. Then next generation of cellular phones may be able to work at 2 million bits, so we can connect well on the net. When I went to Geneva last year, I was surprised by a digital camera  with a radio that transmitted the pictures to a web site and that website could be the framework for the grandparents we mentioned above!

The next slide shows us different radio communication technologies: satellites, multimedia distribution system, all of these have the potential to operate at hundreds of thousands of bits per second, they could be part of the Internet enabled video. This is called ‘Blue tooth’, a new technology that allows devices that are a few feet apart can communicate with each other instead of having wires everywhere, so it is a kind of desktop communications network. The blue tooth radios are pretty small, some people suggest that we could wear them as cufflinks so that ‘at last the right hand knows what the left hand is doing and the other way round.’

... every single nanomolecule could have its own address.

When you’re on the net by radio, you also have to have an address, so it means the address space is being consumed even more quickly because of the use of radio and that is forcing us to look at alternative. One is the 128-bit version 6 protocol. Before we had the 32-bit protocol, version 4. What happened to version 5? The answer is it was an experiment that didn’t work. 128 bits are enough to work with 1038 devices, is big enough to allow every electron in the universe to have its own web page. We made a calculation, we say people live in 30% on the planet, and the habitat that is about 1 km high, if you take that human habitat and divide it by 1038 the volume it represents 46,000 molecules. Do you know the word nanotechnology or nanoengines, the people that works with these nanomachines do not make it in smaller quantities than 46,000 molecules. Then, every single nanomolecule could have its own address.

To summarize, first, the Internet protocol will be literally everywhere in this century, we will see more and more of what is called IP and most of it will be wireless.

... it doesn’t mean that the old media go away, they will be augmented and carried through the Internet.

Second, we will put everything on the Internet protocol: television, radio, telephone, it doesn’t mean that the old media go away, they will be augmented and carried through the Internet.

I think it’s going to be an added value in the services offered by Internet because we will consider it as a place to provide products. Finally, there is going to be a great level of international competition in one dimension at least in the label markets. The best way to illustrate that is an offer I got by email from India, the email offered me to make webpages, 10 pages for 125$, which is actually a good price for making web pages. If I didn’t have a web server to put the information up they rent me a space for 250$ per year. Those numbers are pretty low compared to what we often have to pay in the US. I think we are going to see more people don’t have to leave their own country to work.

That brings us finally to the most important part of our talk: in order to preserve the privacy on the net, we need to encrypt, and this is a way of exporting. It’s really important not only for the business transactions, trademarks, and copyright issues. Probably most of you have listened to music coming out of the computer. That sound is based on MP3 of digitized sound that allows downloading music on the Internet. I actually don’t know if I’m violating the law when listening to this MP3 music, that is why these issues are so important now they open questions about regulations in the Internet environment and much complicated by all the media on the net.

When we carry out transactions on the Internet if there is a disagreement over the transaction, no one is clear who, where and how will this dispute be resolved.

When we carry out transactions on the Internet if there is a disagreement over the transaction, no one is clear who, where and how will this dispute be resolved. So we will have to come up with some decisions . Taxation is also a complex topic. Some people think nothing on the net should be taxed. I think there should not be a disparity between buying in the Internet or in the real world, the two should have equal taxes.

Censorship is a big issue, there are some parts of the world that want to limit the content that is put on the net but what happens if you try to do that? So far, none of the proposals have worked.

Somehow we need to have the equivalent of the wet signature in the digital form. The technology exists already the problem is that we don’t have the infrastructure to implement digital signatures in a global scale. Part of the challenges associated with policy issues on the Internet is that they have to be implemented not only in a national scale but also on an international scale in order to be effective. And that raises a great many questions about how the international agreement is made and so on.

There are a number of problems related to intellectual property. If you download software from the net, whose software is it? is it yours? what happens if it damages your machine? can you complain to someone? What about web content? Is it public domain? Is it protected? what happens if you point to some data on the net and use it? are you violating anyone’s copyright? These are problems that have not been resolved. If someone writes a piece of software that allows you to create webpages, and you use it to create webpages and draw content from another part of the web,

there’s this question of who owns rights in the resulting work?

I would like to invite some discussion and debate. If you are interested in using these slides you’re more than welcome to do it from the address we will give you, they are not protected.

QUESTION AND ANSWER SESSION

(assistant)

The first question I would like to ask is that in these days that we have been overattacked by hackers, viruses, and the vulnerability of the net in general, the media should put their business in other kind of social communication. I would like to start this discussion commenting on what we need first, cops to catch the people that write viruses, or ciberfireman to solve the problems of these viruses. What should we do, all of us, engineers, policy makers and users regarding the level of security we must adopt.

(Vinton Cerf)

In the course of discussing this problem with president Clinton one of the participants said that cyber cops might be the wrong idea and maybe a cyber fire department was a better model. Regarding the notion of cyber cops, there is a kind of open question about what is illegal on the net, that it is not simple to regulate cops and controls because it is not clear what is illegal or not and to what extent.

This is not an easy problem. First of all, during the recent past there have been a number of attacks to the Internet, many of the servers have been overloaded, by deliberately sending too much traffic to them. That is an extremely hard attack to defend against because in some cases you can’t tell the difference between deliberate overloading traffic and legitimate traffic. In the course of discussing this problem with president Clinton one of the participants said that cyber cops might be the wrong idea and maybe a cyber fire department was a better model. Regarding the notion of cyber cops, there is a kind of open question about what is illegal on the net, that it is not simple to regulate cops and controls because it is not clear what is illegal or not and to what extent.

Suppose that in Spain it is decided that visiting someone’s web is illegal, a law is passed, but how can we enforce this law? Now we have an interesting question about the meaning of the law that was passed here, for example that it is illegal for someone to overload a website in Spain. Does that law apply to someone who doesn’t live in Spain? Does it apply to someone who uses the computer in the US?

Trying to create and apply laws won’t work unless there is an amount of uniformity of legislation all around the world that makes the same acts illegal in the entire world.

I think you can see that it would be quite difficult to make that Spanish law applies to someone who is not within the geophysical jurisdiction. Trying to create and apply laws won’t work unless there is an amount of uniformity of legislation all around the world that makes the same acts illegal in the entire world.

I also want to answer Amadeu’s technical question: Can we do something to resist the virus attacks and so on. We have some technologies that will help us resist these attacks because we can identify them but only while the attack is being made. For viruses, the problem is not the network but computers like this one (he points at a computer in front of him). These machines are often built with the idea that there is only one user and therefore there is the presumption that the user of the machine won’t attack his own computer. The problem is that when you put it on the network is accessible to a lot of people and its architecture implies that who is using software. The problem is to build this software that should be much more defensive and much more protective.

Regarding Amadeu’s question of enforcement, it is not clear to me that you can have effective enforcement even with any technological tool. So, let me ask you how uniform do they have to be, can we have laws compatible between countries so that we have some reason to think that things are illegal in country A and equally illegal in countries B, C, and D.

(moderator)

We have only about a quarter of an hour left. You are very welcome to ask further questions to Vinton Cerf.

I am Imma Rodríguez, professor at the Department of Economics and Business Studies at the UOC. First of all, I’d like to congratulate you for this honorary doctorate for your leading role in the creation and development of the Internet. At the same time, I would like to thank you for such an interesting presentation you have made today at our university and to underline its special significance for all of us.

I would like to ask you two questions: first, security problems have been considered one of the main factors that could affect the development of electronic commerce. In spite of this, don’t you feel that in commercialization of products, specially of tangible products there are important barriers to the development of electronic commerce such as the fact that consumers and sellers are not used to this selling and specially in the extra effort required in the distribution.

Second, many analysts consider that the Internet is contributing to the reduction to a number of intermediaries, because with Internet the producer has the opportunity to communicate directly with the final consumer. Do you think that intermediaries such as distributors will disappear only if they are not able to integrate into this system and that the Internet will create other kinds of intermediaries such as those who are specialized in providing information? Thank you very much.

(Vinton Cerf)

Thank you very much indeed for those questions.

With respect to fraud and abuse of the net, this is an area that does deserve legislation

In fact, I would suggest that we have laws today in many countries if not in all of them that say that using infrastructure for fraud is illegal, for instance in the US. I don’t know if the same thing is true here. One of the problems we have is if the person committing the fraud is not in the jurisdiction where the law was passed, for example if you live in the Caiman Islands, where there don’t seem to be many laws, my impression is that it is a fairly open territory. If you commit a fraud from there, something that is illegal in Spain but not in the Caiman Islands, it is not clear how to enforce that. It means that international organizations like the World’s Intellectual Property Organization or the World Trade Organization will have to use these tools to create some kind of community and compatibility in order to protect citizens. I do not know of any technical device that prevents people from committing a crime.

Internet was created to improve customer service, to bring companies closer to customers, but this cuts out many intermediaries, this is what is called “desintermediation”. On the other hand, at the same time, Internet is creating an opportunity for what I would like to call “remediation”, that is to say new intermediaries to add value to transactions on the net.

As for the second question, I think it’s fascinating, Internet was created to improve customer service, to bring companies closer to customers, but this cuts out many intermediaries, this is what is called “desintermediation”. On the other hand, at the same time, Internet is creating an opportunity for what I would like to call “remediation”, that is to say new intermediaries to add value to transactions on the net.

Let us suppose that Manel and I are going to have a commercial exchange, he has something to sell and I want to buy it, but neither one of us trust the other, I don’t know if he’s going to deliver and he does not know if I am going to pay. So, we are carrying out this transaction in our personal computers and we are going to exchange a document that may be encoded describing the transaction. Let us suppose for a moment that Amadeu creates some software that is able to interpret this transcription and that Manel and I, both looking at our web page check the little box saying we want this transaction to be ensured. Amadeu’s software is now activated and provides insurance for both the sales transaction and the purchase transaction. You can even imagine that many other people want to provide this service as well. So, Manel and I check the insurance box and advertising can go on the net saying. How many companies would like to pay to have that insurance! Let us suppose that Amadeu’s software provides the best bid, we will pay him a small fee for insuring the transaction.

His software has added value to the basic exchange. That is an example of “remediating” to introduce new business opportunities into the existing network.

The same thing would be true for the travel business, they will be a big beneficiary of all the activity on the net because through the net we discover new places and people that we want to visit. You can now buy tickets on the net but people still need help to know where the best places to go are, how to plan the vacation and none of that is available from people supplying the tickets. So this service is an opportunity of “remediating” and adding value into the net.

(moderator)

More questions, please. I am afraid we have not much time left.

(assistant)

First of all, congratulations for your presentation. I work in Tarragona, my name is Manel Matés and I am a professor in ESADE. Five years ago people talked about how to reach the mass fast. Today, as you said in your presentation, there are some electronic devices and some new technologies that allow reaching that final user in a very fast way. But nowadays the final user is not accessing the Internet in a fast way, and he wonders if it’s a problem of the computer, the server, the company in Spain, and when you call your server in Spain ISP he tells you that the problem is in the backbone, then, my question is, what has to be done? Is it a problem of MCI? Telefónica? Forgetting about traditional infrastructures and facing the new infrastructures led by the government, should not the government make a bigger effort so that the final user gets this fast access? How can we face the future in order to improve the new technologies to the final user in a fast way?

(Vinton Cerf)

I think that the fastest way to get a high bandwidth would be to create a highly regulatory structure.

I think that the fastest way to get a high bandwidth would be to create a highly regulatory structure. We are finding that in the US today there is even not enough competition to reach high bandwidth everywhere. Cable companies at the moment in the US have monopoly distribution rights and that is not leading to having as much competition as we would like. There are alternative technologies such as digital subscriber loops, point to point radio lines and other solutions but depending on where you are you can not choose because physical limitations sometimes get on the way. So the best radio link may not work, you may be too far away from a central office to have a digital subscriber loop service, and so on. So, the remedy that I propose - which, by the way, is not necessarily being accepted by our federal communications commission - is to have competition within each one of the media. For example, if we have a ‘twisted pair’, which is what we use for our telephone lines in the US, it hasn’t been a 100% effective, sometimes the local provider has not been very cooperative in providing access to his local loops even if legally they have to.

The company I work for, Worldcom, has an important commitment, to open the MMDS radio links to competitors so that it can also reach customers directly. And of course we expect to be compensated for a reasonable amount for providing that access since we provide this radio link. That is the only way I can see to create a diversity of alternatives and at the same time create competitive incentives because particularly incentives to compete are going to be very hard to achieve.

(attendant)

I just want to add a comment, Internet is global and its services are also global, there is only one thing that is typically local: cables. When we speak about wireless Internet we have to accept that this system still has not a reasonable velocity. Parts of that problem should be solved locally. We still have not the competitive access to the local telephone service that is the basis for Internet connection. Some weeks ago, it was announced that this problem is going to be solved in the coming months but we know it will be delayed. A competitive access could be possible, but local monopolies have a price structure, a very low connection fee that discourages from entering this market. We are losing million of pesetas just because we discourage competition in the basic market, which is also true for the international calls, mobile calls, and so on.

Third, you need alternatives of third generation telephones, next year and that is very interesting all the big companies has got a license for this high speed new generation telephony but nobody notices that licenses for radio mobile telephony is something that today could break the local monopoly were granted to the small players so the ones do not have the money infrastructure to really compete with Telefónica. They are doing everything to delay competition to delay these services because Telefónica is investing in Latin America and Lycos and we are supposed to pay the bill.

(moderator)

-  We are now going to ask two further questions to Mr. Vinton Cerf and he shall answer them both. Thank-you.

(attendant)

- Good morning, I'm  Mr. Joaquín Ríos, consultant for the UOC. My first question is focused on the transition matter, moving from one element to another. We have known what the effect 2000 is,  but when and how will the  transition from the version 4 IP to the version 6 be completed?. and my second query would be the following:  what is your prospect over  Governments, such as Spaniard or North American, which take advantage of any  controversy set upon Internet attacks, to take control of the Net? I believe the Net is prepared enough to cope with these problems itself. Anti-virus manufacturers, inter-navigators communication networks,  

are there to prevent from these problems, besides, for instance, we all have been enforced to keep a safe copy of any of our PC 's activity, just in case. To what extent are Governments taking this issue as an excuse to interfere  in the Net?

I have read many comments you have made on democratization of the Internet and I have also heard President Clinton’s comments about this. Don’t you think this democratization is somehow difficult to link with the importance and interest of big companies in the development of the net? Don’t you think politicians should be more on the net? Anyway, I suppose big companies would have to rule this global  democracy.

(Vinton Cerf)

On the IP.4 and IP.6 project I have a team of engineers now who have been running IPV for the last two years and are improving version 6. They are going to implement the transaction strategies they have been exploring and recommending. Before the end of the year I will know if any of these would be sufficient. It is not going to be an easy transition, because it is like changing the engines while the airplane is still flying. There are some devices that most of us called Network translation devices that interfere with the net.

However, these devices are probably the stepping stones to going from version 4 to version 6. For a long time version 4 will be the core of the network and we will have to translate from the version 6 at the edge into version 4 for transport. There will be a lot of version 6 devices that we will use on the net such as in mobile telephones. This is not going to be easy but it is not magic either, it is straightforward engineering.

With respect to government and the potential abuse of the hacker attacks as an excuse to try to control de network, first of all I think that industry has a primary agreement for developing tools. This is a lot like the fire department that can teach you how to protect yourself, how to build homes so they are safer and so on. This does not eliminate the need for a police force, to prosecute people who abuse the network but I think the issue of the industry leadership is very important. If the government tries to use legislation to control things, it may not work because we don’t have a uniform legislation but even more important, technology is changing very fast and it takes time to put laws and by the time you pass the laws they may no longer be relevant.

... it is not likely that the big companies can dominate the network and its evolution at least in political terms, ...

... any company or government that thinks that can dominate the network is going to be surprised just to find that it cannot do ...

... on the net nobody will be prevented from speaking freely. Governments will try to control what is said so, we will all have the freedom to speak but there will be limits on the freedom to what is said on the net.

As for democratization, it is not likely that the big companies can dominate the network and its evolution, at least in political terms, and the reason for this is that Internet still continues to be the world’s largest megaphone. It is a way to listen to voices that would never otherwise be heard. It’s like to say that the true will make its way on the net, because a lot of information will be on the net but there is also a lot of misinformation, and the answer to this misinformation or to information that is not adequate is always more information. So, any company or government that thinks that can dominate the network is going to be surprised just to find that it cannot do that because the Internet has absolutely no concept of national boundaries, it was designed to ignore them. Remember that the Internet was designed for military purposes to work in any country. This system is absolutely ignorant of any national boundaries and as a result it allows always free communication. The best description I have ever heard about how the Internet works with respect to censorship is from John Guilmore, who is the number 5 employee of Sun Microsystems and he described it this way, ‘Internet interprets that censorship is damage and then creates a loop around it.’ I think it is a good definition and I also think we need more control on the net, if we can speak of control at all. At the same time, on the net nobody will be prevented from speaking freely. Governments will try to control what is said so, we will all have the freedom to speak but there will be limits on the freedom to what is said on the net.

(moderator)

- I would like to thank on behalf of the UOC Mr. Vinton Cerf´s for being here today. It has been a great chance for all of us to listen to you. Thank- you everybody, thank-you to our guest and the people who have organized this event.