The Creation of a New Platform for Network Centered Innovation

Darkstrand Plugs Large Enterprises into Innovation As it Teaches Them to Utilize Bandwidth

August 30, Ewing, NJ Darkstrand is undertaking a threefold process (1) to acquire a real national network built out from the NLR backbone by means of sharing agreements with the regional research networks and the state networks.
(2.) to build a network operations staff for a virtual independent facilities based national network and to start operating that. (3) finally and the newest, most difficult, and most fascinating undertaking is to build a company where state- of-the-art high speed optical networking is incidental to a business model that says to large corporations: "we are selling you the ability to initiate research collaborations in 60 days that could have taken 600 days before our arrival on the scene." 


Executive Summary

Needed New Business Models

Telecommunications is no longer about connecting people to have a conversation or to exchange video or data files and be billed for each of those transactions. That is the obsolete business model still embraced by the local phone companies. As the world of fiber optics and the IP protocol has cut the cost of these transactions by orders of magnitude, new business models where the telecom company uses its connectivity services as a business enabling platform have emerged.

In these, the telecom company is used to assist some business processes where the individual communications are just incidental by-products of the larger scheme of things. In this world the telco provides a communications services platform designed to facilitate a more specialized marketplace.

We examine Martin Geddes’ elaboration of this concept where the two sided services platform with the telco as audience builder, matchmaker or cost sharing network may be joined in this case by the new company called Darkstrand acting as an example of a fourth type of more specialized platform - one that exists for the purpose of furthering innovation. However, in this case, we must note that Darkstrand is a more narrow example of the use of a high end optical broadband service that functions as a platform for transfer of innovation from university to large enterprise. As such it will not solve home and small business last mile problems.

We include an abbreviated version of George Gilder’s famous December 1992 fiber sphere essay that explains the problems evident in 1992 that now have created Darkstrand’s business case.

Darkstrand p. 5

Michael Stein, CEO explains how his need for sending large media files for rendering in the late 1990s and early 2000s was not well met by the incumbent telcos. Frustrated, he began to look at the need for tech transfer from the university and research community to the world of large enterprises. Their charters called for tech transfers. Yet not a lot happened and what did get “transferred” was very slow. Why?

The early problem that federally financed research faced in the 90s was that the telco transmission network was the only thing there that could serve as a platform on which to build projects. This was the case despite the fact that the project’s success was not conducive to the on-going business model of the phone company. In a sense it was not until the emergence of Level 3 as an all-optical carrier that a possible an independent alternative to the basic telco network emerged.

NLR, by having a 20 year IRU, for all practical purposes, owned an independent virtual telco plant that could used for goals independent of the old telco business model. That independent plant, it could be argued, made possible a real alternative and this alternative created the opportunity where Stein on behalf of Darkstrand could step in and buy half the capacity of National Lambda Rail giving himself thereby a virtual and independent facilities based platform – something that no one else had.
The possession of Darkstrand’s platform then becomes the foundation that is leveraged to make possible a business model where the platform itself can be both the foundation for, but incidental to, a business based on a goal of enabling technology transfer from universities and national labs to the commercial sector.

The process that Darkstrand therefore is undertaking is threefold:

(1) to acquire a real national network built out from the NLR backbone by means of sharing agreements with the regional research networks and the state networks

(2.) to build a network operations staff for a virtual independent facilities based national network and to start operating that

(3) finally and the newest, most difficult, and most fascinating undertaking is to build a company where state- of-the-art high speed optical networking is incidental to a business model that says to large corporations we are selling you the ability to initiate research collaborations in 60 days that could have taken 600 days before our arrival on the scene.

Darkstrand does this by creating its own process of enabling corporate customer to identify where the talent is that their engineering staff needs and then facilitating the process of statements-of-work and contracts and finally offering project management assistance to bring in ongoing projects on time and on budget.

Darkstrand is talking about building a National Innovation Network – a worthy and important subject. It is something that they can do because their unique circumstances put them in a position where their network is essentially a utility rather than perceived as the focal point from where all their profits must come.

The execution will call for the simultaneous collaboration and cooperation of many different people with different levels of skill. But a point of major significance is that Darkstrand is treating high speed optical connectivity as the general utility that we must have for a productive future. It is something like the electric grid that is just there and you use when you need it. It does not presume like ATT to be able to make business records out of its customer’s data traffic.

Darkstrand is moving ahead of four fronts. First an MOU with Florida Lambda Rail that Mike hopes can be used as a template for last mile agreements with the more than a dozen other regional optical networks. These MOUs will extend the Darkstrand footprint from the backbone into hundreds of cities. Second is an MOU being negotiated with the Illinois state network where it is hoped that this MOU can be presented to many other state networks that can benefit from connecting to the infrastructure that Darkstrand is trying to build out. In connecting corporate and research America Mike says he would like for there to be a last block problem rather than a last mile problem.

The third front is the creation of series of alliances and related MOUs and contracts that will facilitate Darkstrand’s ability to inventory university and laboratory technology incubators and tailor contracts that will facilitate the ability of corporate customer to identify sources of innovation and create collaborative efforts that by beginning new work projects will enable them to create projects and bring technology transfer to their respective markets in record time. And the fourth and final front of course is the recruitment of corporate customers that Darkstrand would like to sign up for 10 gig wavelengths at $100,000 a month for a period of 15 years.

Symposium Discussion

New Low Carbon Notebooks, p. 28

St. Arnaud: a good article on how I think the PC industry will evolve in order for consumers to reduce their carbon footprint.

Jaap van Till: What now is happening is that the left side (in horizontal value chain terms) trend towards ever smarter PCs software and hardware is disrupted too. Functions are moving leftward into smaller & less costly Netbooks, Smartphones, Net-Tops AND rightward towards even bigger "online SAAS services for processing and storage like Google Docs etc. Much to the chagrin of PC software maker Microsoft and the makers of PC's like Dell. In my opinion the desktop PC is taken apart in the appearing vacuum by this trend into a "STUPID PC".

John Levine: It's the Wheel of Reincarnation that Ivan Sutherland described in a paper in 1968. Users have specialized devices (in his instance graphics terminals) that grow so many extra features that they compete with the larger system to which they're connected, so the device and the system merge until the next specialized device shows up. (This time the next device will be your phone, of course.)

Telcos Harming the National Interest, p. 34

Vint Cerf: In the places where there is a strong regulatory control, it seems to be working. Although if you are watching the UK the BT guys are making the same kinds of arguments that Verizon ATT and the cable companies are making. Basically it's like little kids throwing a tantrum - I am not going to build this system unless you give me three scoops of ice cream and a pony.

My reaction to this is quite negative. It is harmful to the national interest to behave in this way. Because this is serious infrastructure. It is very much like the road system and I have been ridiculed for suggesting that it be treated like the road system.

"Oh you want the government to take over."

No. What I really want is a split in the regulatory framework for Internet service. I want to see a reintroduction of common carrier responsibilities. I want to see a horizontal treatment of regulation. I want to see the broadband providers split into two parts either literally or at least from the accounting point of view. I want wholesale broadband service to be required. I want them not to be able to interfere with anyone's applications. I want them to charge themselves the same for their access as they charge everyone else. All of those things. And they will say well we won't build any more fiber.

I am not an economist and I do not know what the right reactions are other than to say we have to derive incentives that will cause these companies to behave differently or to create an incentive for a competitor to put in facilities that will compete with them. [snip] So we have to have a set of rules that makes sense out of attempts to build multiple infrastructure and in the absence of this we have to make it a privilege to build the infrastructure. There has to be a reasonable rate of return, but there cannot be a confiscatory rate of return and it can't be abused by allowing people to throttle competitors."

Tim Cowen: The issue now is where next? Turn the clock back and seek structural separation or move with the times and adopt the functionally separate route that is working and would be capable of being implemented in the US much, much more quickly?

Ireland, Sweden, Italy, Australia, and a number of others are looking at functional separation as an effective remedy and it is supported by the European Commission as a remedy that should be available to national regulators as part of their national toolkit. Why not in the US?

Peak Hierarchy p. 37

In the 21st century bigger is not always better and former economies of scale become diseconomies.

Bauwens: As you perhaps know, I have been more and more interested in the age of the transition to feudalism, where the Roman leadership similarly could not understand the new christian mentality (see

I think there is an important issue here: do distributed modes really outperform the centralized/decentralized modalities. If this is indeed so, then I suggest that we introduce a new concept: Peak Hierarchy, which I'm using as an analogy with peak oil, see

Cost of Service and Net Neutrality p. 38

Werbach: OK, so we've put to rest the nonsense that non-discrimination is a recent invention of The Big Bad Google. Here's the more important question -- how does the history of common carriage that Tim Cowen and Bob Cannon relate help today's policy-makers decide what to do?

Cannon: Or, like so much jurisprudence, it can improve through trial and error - as the FCC improved from Computer I to Computer II - noting that you cannot improve until you have taken the first step. American jurisprudence is based on competition in the marketplace of legal ideas. It assumes that the first answer is not necessarily the best answer - but that we will not get to the best answer until we vigorously consider the issue and conflict and try a few times to get it right. If we fear that we will not be perfect the first time, then we can never act. American jurisprudence is premised on the idea that we need a few good tries, in lieu of perfect tries, before we get it right.

I am not taking a position on whatever it is that happens Friday - only taking a position on the phobic position of doing nothing less we get it wrong.

AMS-IX Slowdown p. 40

Odlyzko: "Network neutrality, search neutrality, and the never-ending conflict between efficiency and fairness in markets," to appear in Review of Network Economics, [snip] In particular, and perhaps Hendrik or Jaap on this list can shed some light on the issue, the Amsterdam Internet Exchange, AMS-IX, has its traffic growing quite a bit more slowly than in previous years:

van Till: Yes, there has been a slowdown in the growth of traffic through the AMS-IX, and Rudolf mentioned a number of probable causes, but that does not mean that TOTAL internet/telecom/video traffic growth is slowing down. snip

A real stampede seems to happen on mobile data! Hendrik told me that several Telco's in Europe have reported triple digit growth/quarter in data flows on mobile devices. If you remember my suggestion, this can be written as W(mobile data)= 3 months, a real explosion in demand by the hyperconnected! This is in line with my "Networked Stupid PC" trend. Don't get trampled under the hooves.

Permission Problem p. 42

I personally think that the very fact that it is hardly possible to get a real software patent on a mathematical algorithm is one of the very reasons why software writing has ended up stuck in its situation of being mainly a cottage industry and has defied any serious attempt to industrialise and get serious laboratories operational. The copyrighting in software is only a proxy for what software and computer science really should be about: the invention of smarter algorithms and bundles of algorithms.

Goldstein: The US patent system can't be fixed all that easily. It lacks trained people, and its pay scales aren't competitive. The employees are rated on productivity: How many patents did they approve today? That's theoretically fixable, but the reality is that serious reform, to require serious searches of prior art (not just prior patents) and serious evaluation - of overall patentability, is not going to happen within decades. The system is too broken. Yes, it's the US in its usual self-destruct mode, leaving everything to be settled by the litigators. And we have this abomination called the Eastern District of Texas where the courts have literally become a local industry.

Regulatory implications of FiOS p. 48

Cerf: I have been suggesting a telecom re-write that would place Internet a new title, not Information service, not traditional telecom and not cable, but rather, a new Internet title. It would have common carriage obligations, neutrality requirements, etc. Am I on weed?

Cecil: Either that or the FCC is smoking the same bud. :-). Seriously, though, hasn't the FCC already achieved more or less the same thing informally? More deeply, what concerns me about any incremental change to the 1934 act is framing. We would continue to amend the same sausage mill expecting different results. I don’t think it is possible anymore for many reasons but perhaps the biggest is that the exceptions have swallowed the rules.

Cerf: They haven't - they've wrecked common carriage through use of title I for Internet.



Needed a Business Model for the New Economy While Seeing an Architecture of Opportunity

Darkstrand as One of Several Possible Alternatives


Understanding the More General Model p. 1

Darkstrand Has a Fourth Type of “Platform” p. 2

Mandatory; a New Business Model p. 3

Darkstrand Plugs Large Enterprises into Innovation As it Teaches Them to Utilize Bandwidth

A Conversation with Michael Stein CEO p. 5

Building Bridges to the Research and State Economic Development Sectors p. 7

Doing the “Deal” p. 9

Operational and Cost Issues for Enterprise Customers p. 10

Expanding the Darkstrand Footprint p. 12

MOU, Interconnection and Operation – Florida and the Other RONs p. 13

The State Government Networks p. 16

Chicago and Other State Networks p. 17

Enterprises to Use Both High Speed IP Network and Innovation Platform p. 18

The National Innovation Network p. 20

The Contracting Approach p. 23

Transectoral Contracting, Collaboration, and Speed to Market p. 23

The Scenario p. 25

Program Managers to Guide Projects p. 26

Conclusion - Stages of the Business Model p. 27


Symposium Discussion July 17 - August 17 2008

New Low Carbon Notebooks to Connect to the Internet p. 28

Vint Cerf - the Telcos are Harming the National Interest p. 33

Peak Hierarchy p. 37

Cost of Service and Net Neutrality p. 39

AMS-IX Slowdown p. 40

Patents and The Permission Problem p. 42

The Regulatory Implications of FiOS Technology p. 48

Executive Summary p. 55