Vint Cerf on video distribution Video

Vint Cerf on video distribution Video Transcript

>> So Vint from the old days of the creation of the Internet to now, what's happened and sort of, what's surprising to you. Tell us a little bit how you see the evolution of the Internet.

>> Well, Bob Kahn and I of course are often asked if we imagined what we have now and the honest answer is not exactly. But we actually did have a lot of experience with many of the applications that are common today way back then in 1973 and even before that. We built a lot of this on top of the experience with the ARPANET, which was funded by the US Defense Advanced Research Project Agency. So where we are today is about 20% penetrated, about 1.3 billion users, so the Chief Internet Evangelist at Google -- that's me has another 80% of the world to try to convert. We need a lot more Internet access all around the world and we need a lot broader bandwidth access to the Internet for everyone. And the reason you and I are talking today is there's an initiative called Internet for Everyone, which is intended to highlight the need to have broadband facilities everywhere, not only in the US but elsewhere, especially in the rural parts of the country where often there is little or no broadband access. The reason that broadband is so important is that it enables a wide-range of applications. The Internet was designed with no particular application in mind, in fact and that was important. It simply carries pockets from one place to another. And that simple service -- on top of that simple service, virtually anything can be build, so what do you see today. You see web pages, you see blogging, you see videos uploading and downloading, you see people chatting and twittering and IM and what not. You see all kinds of applications that are more or less independent of the way the network functions. All they do is use their pocket-carrying capability. So, we're worried frankly, in the United States especially that we are inhibiting our own innovation and creativity. We are not creating jobs that we could create if there was broadband access everywhere in the country.

>> Could that be solved by WiMax or by cabling or by TelCos. How could that be solved in next ten years.

>> Well, the real answer is that there's no one particular technology that necessarily is the best solution everywhere. The highest piece solution is tend to be fibre, but there are places where broadband wireless may turn out to be the most cost-effective way of delivering high-speed or even satellite, although, synchronous satellites have a delayed component in [inaudible], a quarter of a second, up and down because of their location. Medium Earth orbiting satellites on the other hand have a very small delay -- 30 nanoseconds or so, in which case that might be another potential kind of solution. What I believe we're looking for, certainly I am anyway, is to find as many different broadband solutions as possible, so that they can be evaluated for their cause and deployed in places where they're most effective.

>> You know, I spoke to Bob Metcalfe and he told me that he believes that online video in a sense is environmentally correct or that it saves energy. Could you talk about that a little bit?

>> Bob and I haven't discussed it, but I think we probably would agree that you can substitute telecommunications for transportation. It's a little tougher if the time zones are sufficiently diverse, but for people who are a few time zones different, that actually works pretty well and with high enough bandwidth again, you can get a very good quality interactions. Companies like Cisco sell a Telepresence capability, but it's beautiful and it's wonderful and it's life-size and it's expensive.

>> Yeah.

>> It's about $300,000 to outfit a room. Those costs will come down with time and I agree with Bob that you could substitute telecom for transportation. But there's something else we can do too. We can use information technology to calculate our use of the petroleum products that we need by optimizing the routes that we drive or fly or ship. And so, here's another opportunity for computers to contribute by making more efficient use of the available resources.

>> One last question, you know, we are covering video in a big way, did you anticipate that video would be a big part of the Internet and how does that challenge the tubes and how could that sort of grow or and bandwidth. We're talking about these big files in particular.

>> Let's start out by observing that many of us were doing experiments with video over the Internet in the late 1970s and early 1980s along with the voice over the Internet. So, that wasn't new, but the quantity was new. The Internet has scaled up substantially. All those scaling is now its biggest threat. We need new address space and that's why IP Version6 is important. What I'm foreseeing frankly is that video will be used in download mode more than it will be used in streaming mode as time goes on. If the capacities of the net get high enough, you can download video, you know like a gigabit per second would let you download an hour's worth of video in 16 seconds. Kind of like what happens with the iPod where you download music faster than you could listen to it. So I anticipate that a lot of video, the people who watch will have been downloaded and then played back whenever they want it in sort of TiVo style. Perhaps even more important though, Internet is not taking advantage of broadcast media. We turn broadcast media into point-to-point lengths with what you see with Wi-Fi for example. What we could be doing is rethinking some of the protocols, so that if you have a broadcast medium, you actually use it to deliver the same thing to a large number of people at the same time, which what television and radio do. It's very efficient, it's very effective. We're delivering the same thing, so 65,000 people who want a particular movie then you broadcast it. Don't send 65,000 separate copies. If two people want a copy, it might be perfectly sensible to just to deliver that to them as a file transfer and let them play it back. The cost of storage is so low, the cost of digital processing is so low that doing things in real-time may not be necessary anymore. And so I'm anticipating a shift in the way people use video and also expecting to see video as an interactive medium, which means that advertising in that medium will have to change. Instead of being something which is forced on you just at the moment when the scene gets all exciting. Instead I think people would want what they have now in the Google world where they can decide what advertisement to watch or not. I think they're gonna want the same kind of control in the video medium and it's possible to give it to them.

>> You know, next year, we'll have a new President. We'll have a new Congress, what should be on the top of the agenda?

>> Well the principal one is getting policies together that will cause more broadband to be rolled out everywhere in the country and that means providing incentives for people to invest in the infrastructure. It may actually mean a regulation of some aspect of the Internet in order to assure that there is either competition or fair access to underlying broadband resources. In some countries for example, the broadband provider is required to provide a wholesale access to broadband services, so as to allow multiple parties to compete with higher level applications, which may in fact produce a larger revenue stream than the basic bit carrying. So that's worked out pretty well in England, it's worked out in Japan, it's worked out in New Zealand and so on. The U.S. is far behind in terms of its regulatory posture. It's still very hands off. And although that's tempting, I think as a nostrum, it does not worked out very well. So that is a big issue as a broadband capability. The other one is getting IP Version 6 into the Internet, so we don't run out of address base. And after that it's all a question of people inventing new applications.