Forget everything you know about the Internet
This morning I visited the CSAIL at MIT and had the pleasure to discuss the future of the Internet with Dr David Clark and Dr William Lehr. Both are engaged in the interesting Communications Futures Policy project. The subject of the day was the future of the Internet and the conversation keyed in closely with the conversation I hade just a few days ago with Dr Parulkar.
What technological and social trends will drive the change for the Internet?
Both Dr Clarke and Dr Lehr agree that the Internet will change because of new technologies and networks that connect to it. Sensor networks and embedded computing are things that will become more and more important. But it is important, notes Clark, to think about this question in two steps: one is how the Internet will change and the other how it should change. The truly hard question is not how we will make the Internet better – we can envision ways to do this, as can industry – but rather how we can create the infrastructure we will need in the future. This is also the good thing with the GENI-project, according to Dr Clarke. The facility is a palpable result of the project but it is only one part, the actual research in the FIND-project (Future INternet Design) is much more important (a facility without projects to run would make no sense).
We have a peculiar problem when it comes to this. The Internet has been so hugely successful that it tends to blot out the need for not only incremental, but rather disruptive, innovation. It is hardly surprising that the majority of Internet researchers work with a better net, instead of a completely different net, since the net works so well right now. But what should the Internet look like in five to ten years time? That should be the question and we should bracket all issues of transition to this new platform, and concentrate on re-inventing a global communication platform as we would envision it not knowing of the Internet of today.
A hard task indeed.
Is the security situation degrading or improving or is it stable?
Both Dr Lehr and Dr Clarke think that the security situation is degrading and that the Internet is becoming less secure. Security, in fact, is the primary reason and motivation behind the FIND-project, Dr Clarke says. Phishing, new security holes and other systemic risks become more and more obvious, and where credit card companies could traditionally internalize risks by becoming their own insurance companies they have a hard time doing that today because of the systemic component of Internet risk: a disruptive technology could very well destroy the entire security systems of the existing commerce infrastructure.
Dr Clarke also thinks that in the next ten years information security will transform into risk management, the ambition to achieve total security or to use encryption will wane and be replaced by a reasoned attitude towards risk.
But what does it actually mean to say that the Internet is becoming less secure. Would it for example be reasonable to assume that information security breach related costs (ISB-costs) have increased as a percentage of the net economic worth of the Internet? Well, this is a much harder question to answer, says Dr Lehr. We do know that information security related costs are increasing, but the again, the total economic value of the Internet is almost immeasurable today and still growing (yes, that does sound strange, but it is huge anyway). It may well be that ISB-costs actually are becoming relatively smaller since the economic potential of the Internet is growing so quickly.
What about end-to-end architectures?
Much as Dr Parulkar Dr Clarke is not comfortable with the question of whether or not we will keep certain aspects of the existing network architecture, such as the end-to-end design. He does believe that a central component dealing that enables parties to freely consent on the use of different applications must be retained in order to ensure innovation at the edges of the network, but he also acknowledges that this is a blunt description.
"We need a new language to describe the richness of what is happening with networks today", he says.
And this new language is something he is trying to figure out in his ”non-existent spare time”. This does not mean, however, that he condones the idea of legislating for network neutrality. The idea or thinking behind network neutrality is that a infrastructure owner should not be able to prioritize his or her own traffic while sub-optimizing that of other, virtual operators of different kinds. Large companies dependent on the network, but unwilling to pay for it such as virtual VoIP-providers, search engines and others have been pushing for legislation that would make it impossible for network providers to discriminate against service providers.
The temptation for these companies to do so, however, must be great. To be able to sell prioritization would be to transform connectivity from a commodity to a differentiated product and would probably lead to enormous profits when media companies, service providers and others had to buy prioritized traffic (or simply traffic of with a certain QoS).
The problem that Dr Lehr and Dr Clarke have with this idea is not that they are against the legislation per se but rather that they think that this is a very dangerous area for the legislator to move in. The slightest mistake may lead to very adverse effects, and they both state that they have a hard time figuring out how this legislation would even look. All current examples the reject outright as not well thought.through.
"I am extremely nervous about our ability to write laws that cleanly excise the bad behaviours", Dr Clarke says.
And what would immediately happen if this was allowed is of course that basic Internet access would get a baseline quality that would be far below today's levels.
In summary
Both Dr David Clarke and Dr William Lehr have a wealth of experience in the Internet field, and their priority right now is to find the next Internet, beyond the current incremental improvements. What really strikes me about the discussion we had is how hard this will be, how extremely challenging it will be to bracket the Internet, it's current successes and the huge societal impact it has had.
And how necessary it is to do just this in order to come up with something truly deserving of the epithet ”the next” big thing.