7:10 pm Second Discussion Panel (introduction by Paul de Sa)

Ajay Agarwal, Managing Director, Bain Capital Ventures

One of the risks for an innovating wireless company is that the innovator has to go to each of the carriers and negotiate an API in order to gain access to network. This increases risk and hesitancy to invest. In the last two years with the advent of the iPhone – it is not totally open – but it is a huge leap forward. Skyhook’s technology is built into the phone to identify location. Have seen an explosion of applications that take advantage of this location based information. Yelp uses it to provide reviews of stores and restaurants nearby. Another example is UrbanSpoon; shake your phone and it will recommend a restaurant nearby. This is a great story of how an open system unleashes innovation. As ecosystems gets more open, will get more investment.

The greater the threat of regulatory uncertainty, the more we hesitate to invest.

Nabeel Hyatt, Founder/CEO, Conduit Labs

Disruptive innovation which poses a threat to other parties; we have to think about how important disruptive innovation is to the overall economy and ensure that it continues to occur.

We developed an application where we had to talk to each of the carriers. It was a new device and new use for the network. It was easy to get Wired and Oprah to talk about it; but could not get executives of carriers interested. Small number of gatekeepers make decisions about what will be on the network. Right now, in this environment, I would not start a mobile company.

Anything but neutrality ends with a small selection of services.

Amy Tykeson, CEO, BendBroadband

I see that I am the only one here actually running one of these networks.

Three major concerns with network neutrality:

Is it broken? The current regulatory environment has fueled investment over the past decade. So how is it broken? Worry about government getting involved in an industry which is so dramatically changing.

Ecosystem which continues to blur. Look what has happened to search and how it influences what people do on the web and where they go on the web.

ISPs have to be allowed to manage their networks to preserve their customers’ experience.

I worry about unnecessary rules that would increase our costs and increase uncertainty over litigation over what terms actually means. Light-handed regulation has enabled our investment to grow.

7:23 pm

Christopher S. Yoo, Professor of Law and Communication, University of Pennsylvania Law School

One of the things that concerns me is the extent to which the policy debate has been framed by a 1990s vision of the network.  The universe of users and applications and the technologies and business relationships comprising the Internet have become much more diverse.The market for the Internet in the United States may be starting to approach saturation . When you reach the flat stage of growth, the nature of competition changes. Instead of offering general products targeted at new users, service providers naturally shift their focus to using more specialized production processes to reduce costs and to providing specialized services that provide greater value to current users.

A number of technologists believe that the Internet is locked into an obsolete technology. The existence of a large installed base creates natural inertia. The hierarchical nature of the current network makes it even harder to evolve to a new protocol stack. While the current architecture promotes innovation in applications that fit within the current protocol stack, it retards innovation in applications that require a different architecture.

The nature of applications is changing. The previous world was dominated by downloads. The current world is increasingly shifting toward interactive video, which is more bandwidth intensive, less tolerant of delay and jitter, and uses protocols that will require new tools to manage congestion. The focus is also shifting from content to applications.

Historically, network providers were able to increase capacity by leveraging legacy investments. Capacity growth now requires more substantial capital investments. Carriers are attempting to use network management to reduce cost of capacity expansion by shifting capital expenses to operating expenses.

We should expect things to change. The current architecture may be inhibiting certain innovation. Lots of things the current Internet does not do well, such as mobility and multicasting.  Other functions are shifting from the edge into the core of the network, such as botnet detection.

7:29 pm
Questions and answers.

Barbara van Schewick: I did not hear anyone saying we have to keep the current architecture of the Internet . What I did hear is that there are certain factors, and these factors give you the framework in which the network can evolve.

Amy Tykeson: Innovation is happening in the network, to improve the experience of the customer. Points to DOCSIS.

Greenstein: Most important thing FCC can do: More Spectrum!

7:47 pm Adjourn

6:00 pm First Discussion Panel (introduction by Sharon Gillett)

David Clark, Senior Research Scientist, MIT Computer Science & Artificial Intelligence Laboratory

QoS is not inherently opposed to innovation. But QoS causes fear. Don’t want fear to dominate hope. QoS was originally intentionally application blind. QoS field is not the same as the application field. User should be able to pick QoS of applications they want.

How do you get QoS in interconnection across Internet; problem of interconnection on Internet. I am not encouraging FCC to regulate interconnection. ISPs have to figure this out but ISPs have concern about antitrust. Need not regulation but facilitation.

Look at all the actors. This is a balancing act to make sure everyone can innovate. There are very few absolutes. The Internet is not absolutely open. It is not absolutely for the benefit of the ISPs.

Need facilitation across actors. Previously done through IETF.

6:06 pm

Tim Berners-Lee, Director, World Wide Web Consortium (W3C)

I invented the Web 20 years ago and it was really nice that I did not have to ask anyone’s permission. The Web was an innovation on top of the Internet. It has become a platform on top of which people innovate. EBay happened on top of it. HTTP also was not designed for a specific application or for a specific company.

I have spoken out previously about packet discrimination. Example is the delivery of packets to the shoe store of choice. Without neutrality, could use discrimination to control what sites you reach to find out information about religious information or information about evolution.

I have benefited very much from the Internet as it originally was, where I did not have to ask anyone for permission. I know that the Internet had little regulation but it was originally done through good will. There are some times when you have to have rules – may they be simple – because they are essential to the way that the thing works. We may need regulation if we cannot assure it without it.

I would not use network neutrality to mean QoS. Neutrality means ISPs cannot use what it knows about me to control what applications I use or content I see. Non discrimination does not mean that you cannot have QoS; but there is a fine line where I cannot download a video because the network service provider is also a cable video service provider.

6:13 pm

Susie Kim Riley, Founder/CTO, Camiant

I agree with Barbara that openness is critical. The ability to try lots of different things and then try them to see what sticks. What has enabled a lot of these innovations to take off is the investments that networks have made in their infrastructure. Investment has enabled access to these applications. If  a network blocks access to applications, then they will lose subscribers. They invest because they want to roll out new services like video conferencing. By doing this, they collapse many networks / services (telephony, video) to one IP platform. Because they invested in this network, it benefited many applications that leverage this platform.

We need openness and transparency because people need to know what their service is and how the network will perform during congestion.

But one aspect of network neutrality has dampened investment. This is the non discrimination requirement that does not allow network services to provision QoS. Some of the applications that operators are talking about differentiated applications where subscribers can pick and choose what applications they want differentiated with QoS. In Europe these type of services are starting to roll out. These services are innovated, give consumers choices and provide differentiated services for consumer. This allows, for instance, a gamer, to differentiate gaming for QoS.

Application and network innovation are not mutually exclusive. Operators cannot keep adding capacity for the same price. With the non discriminatory provisions of network neutrality, you prohibit network innovations in differentiated service.

6:21 pm

Aaron Ahola, Akamai

Openness of Internet has been crucial to success. Akamai came about because the centralization approach of early Internet led to congestion. Originally people thought WWW meant “World Wide Wait”. Akamai solution was to make content more accessible in a more diverse way. Solution was to deploy servers at edge of the network and cache content, along with easy synchronization with original source. Bypass the choke points of the Internet. Now have about 70,000 servers. Today, we accelerate 15-20% of world’s web traffic. What we have seen from our customers is that customers are investing in the web and it is for a couple reasons: common standards, global nature of the Internet, how quickly you can change or adapt things on the web.

6:34 pm

Barbara van Schewick: There is a real need for regulation to draw the line on what the rules are; the uncertainty as to rules has led some to be hesitant to innovate.

6:58 PM Break

FCC Workshop January 13

4:50 pm Panelist Presentations

Barbara van Schewick, Assistance Professor of Law, Stanford Law School

Three Stories: First, EBay, a huge success with 88 million people world wide use EBay

Internet built on design principle called end-to-end (E2E) arguments described Saltzer, Clark and Reed in 1981 [PDF]. Network provides only general services of broad utility across applications; Network not optimized in favor of specific applications; all application-specific functionality at end hosts. Network is general. The Network is application-blind. Innovators decide what they want to build. Users decide what applications they want to use. Low cost for innovation. Many people can innovate.

Second Story: Skype

Third Story: Google Video

Today service providers can control the applications and content on their networks; they may do so in order to increase profits, exclude unwanted content, or manage bandwidth. By singling out specific applications, network providers start picking winners and losers on the Internet. Innovators cant innovate; Investors don’t invest.

Measures that increase the costs of application innovation reduce the number of innovations that will be justified and reduce the size and diversity of the pool of potential innovators. Enabling widespread experimentation at the application-level and enabling users to choose the applications they prefer is at the heart of the mechanism that enables innovation.

Widespread experimentation of applications, between users and innovators. Users and network providers use different criteria for what applications will be successful. Network providers will have different criteria (does this compete with one of my applications, does it consume my bandwidth). Consider Skype. Many mobile providers do not permit consumers to use Skype because it keeps users from using their service.

No body knows if an application will be successful until users try it; network providers can not replicate this. If network service providers pick winners and losers, then users will get applications that they would not have otherwise have chosen.

Low cost of innovation makes many more applications worth pursuing and allow a more diverse group to become innovators.

Why do we care about innovation: it contributes to economic growth and creates value for society. Did you ever try to explain to a grandmother why they should get the Internet? You don’t explain that it sends data packets back in forth. You explain that if you get the Internet, you can get pictures of grandchild and allows you to read about anything. The Internet creates value by allowing us to do the things that we want to do. Applications are the means for allowing us to do these things.

Prof. Shane Greenstein, Elinor and H. Wendell Hobbs Professor, Kellogg School of Management, Northwestern University, Transaction Cost, Transparency, and Innovation for the Internet

Open has many meanings; the meaning in this talk is communications between partners in a value chain and transaction costs: Cost of setting up new process to create value: Cost of executing series of actions within a value chain

The important economic role of transparency and consistency for lowering those transaction costs: Lowering transaction costs supports wide participation in innovative activity.

Internet is the basis for a wide array of services and economic activity

The Legacy: Collective invention is a process in which improvements or experimental findings are regularly shared

It becomes something different when it commercializes. What’s new with the commercial internet: Platform: reconfigurable based on compatible components. The Internet today is a network of platforms. The second thing that has changed is broadband concentration which has resulted in small set of carriers in local markets with market power. This raises a traditional concern about prices.

Transaction Costs: Transparency. Participants know what changes are imminent. Participants actively inform others in advance. Any participant can learn info without restriction. Originally the Internet was quite transparent.

Other Change: Consistency: Processes and policies for interacting with other firms change slowly (at most). Matters because it shapes return on investments (ROI). Entrepreneurs & VCs dislike inconsistency. Opportunistic pricing or renegotiation costs.
Factors shaping ability to capture value. From broadband firms with market power? In developing apps that work with multiple ISPs

What Shapes Participation in Innovation: Processes for soliciting input in decisions Drawbacks: costs of notifying/communicating, strategic costs from giving up control over direction, slowing speed of decision: Gain: additional info. Variety of opinions.

Conjecture: Gains to industry far higher than cost of transparency and consistency. Transparency & consistency increase participation in innovative activity. In a competitive setting, those without transparency wont work with partners and will have trouble in market. Firms with market power, platform providers have weak incentives to be transparent & consistent.

What Shapes (Re)negotiation: Interdependence + changing opportunity means partners must reconsider in time and frequently: Concern about powerful firm using negotiating position to limit what information they give others. Worry about take it or leave it offers, or refusal to deal. Hid behind lack of transparency & lack of consistency.

When all broadband ISPs look less transparent and/or consistent, what then? ISP owns users and applications. Transaction costs for innovation increase. Decline in breadth of participation and in extent of innovation.

Summary

Wide participation in innovative activity encouraged by transparency and consistency: Lowers transaction costs of participants in interdependent value chain. Concern: Incentives of broadband firms w/market power & platform vendors to be transparent or consistent?

If commercial Internet necessitates some active role for management at carriers… Implication of unrestrained discretion? In presence of market power at broadband firms & rise of platform vendors?

Glimmers and Signs of Innovative Health in the Commercial Internet (2009)

Marcus Weldon, Corporate CTO, Alcatel-Lucent, Innovation and Investment in the Internet: Past, present and future

The Internet is a series of interconnected private, managed IP networks and public networks that connect through transit links and peering links. It is Open because any user can connect to any internet application located anywhere on the internet from anywhere else on the internet. The Open Internet supports ‘best effort’ transport, with no service guarantees, but ~equal usage of the resources between users connected via the same access ‘on ramp’. The ‘on-ramps’ to the Open Internet are primarily provided over managed IP networks, which also support managed services with specific QoS guarantees

But not all Internet services are equally accessible:
• Content Data Networks (CDNs) enhance delivery of content
• Networks of Data Centers enhance access to applications
• Different service tiers are offered to improve users’ access

Best Effort Services and Managed Services : What’s the Difference?

Best Effort services : Typically delivered from a web server on the Open Internet to the client, using the Managed IP network as an Internet ‘on ramp’

Managed services: Typically delivered from a web server within the Managed IP network to the client and only traverse the Managed IP network

(Demonstrations of how different services fit within these models, some hybrids – more diverse array of applications is better, giving user ability to chose)

What led to the success of the Open Internet and what harms it?

Investment in Content and Applications – was helped by openness of Internet (key factor), affordable broadband, convergence of networks to IP, simplicity of IP, proliferation of IP-enabled devices – was hampered by bandwidth limitations, limited reach of infrastructure, security and rights concerns, usage of IP-enabled devices, and poor quality of experience.

Investment in network infrastructure – was helped by FTTx-promoting regulation, competition, demand, network convergence, massive IP market – was hampered by FTTx hampering regulations, geographies and topologies, cost of networks, lack of competition.

Open is good; but have to have the ability to afford the bandwidth.

What’s the problem: While network revenue (per subscriber) used to exceed cost, revenue and cost have converged (mobile cost is forecast to exceed revenue). It is bad for everyone for the network to go out of business.

What’s the solution: managed services for revenue sharing and investment for good of all. Allow user to select QoS for the applications that they want; increasing revenue for network. Experience gets better; more applications developed. Even Best Effort network experience will improve.

5:41 pm

Jeffrey Glueck, CEO, Skyfire, Perspective of a Mobile Application Developer and Entrepreneur

skyfire launched in 2009 after three years of development. skyfire is a fast mobile web browser that supports all of the rich media on the web, allowing users to watch any video on the web including Flash, Silverlight, and Real Player, and to visit any website (not just mobile sites). It is free and downloadable from the skyfire website, and runs on Windows Mobile, and Nokia N and E Series phones.

Features of Internet that have given confidence and predictability for Investment: Mass Scale; Interoperability; Low Barriers to Entry; and Adequate Infrastructure. One website would work on any computer (Dell, Apple, IBM). Did not have to be approved to get into marketplace. Knew that photos on website could get through the network.

Mobile Internet presents new challenges: Closed proprietary standards create fragmented markets, lowers ROI per unit of effort. Lack of interoperability

Three obstacles to innovating on the “closed” mobile Internet: (1) Freedom of the consumer to choose apps and defaults; (2) Lack of Transparency and consistency on network management practices, and discriminatory requirement for proprietary protocols; (3) Uncertainty on non-discriminatory and adequate data throughput.

Skyfire has to date held back from developing for iPhone because of cost of development, Apple’s rule against duplicate functionality and opposition to Adobe Flash playing on the iPhone. This is a limit on Skyfire’s ability to improve the browsing experience.

Carriers block certain ports, do not disclose port blocking practices, throttle RTSP streaming, and block applications from their stores. Apple declare video must use HTTP streaming; our app was not designed for HTTP streaming, our protocol is better, but not allowed to use it.

Non discrimination should be hallmark of network management.

Concern: take of allowing biggest players to colo servers inside ISP host; if it is not provided equally to other firms, that means new entrants don’t have a chance because they cannot afford to collocate.

“Mobile Warming.” Browsers provide key interoperability to enable competition. Browsers allow small entrants. Publishers can not be expected to build and maintain apps for each OS/Hardware permutation and version.

4:30 Introduction

• Sharon Gillett, Chief, FCC Wireline Competition Bureau
• Paul de Sa, Chief, FCC Office of Strategic Planning and Policy Analysis

• Chairman Julius Genachowski appearing via video, welcoming participants, thanking FCC staff, and inviting public participation and interaction
• Commissioner Meredith Attwell Baker, present at the Workshop, welcoming participants

4:40pm Framing Remarks

Sally Shipman Wentworth, Internet Society’s Regional Bureau Manager for North America

“What has remained constant about the Internet, what is at the heart of its growth—and what distinguishes the Internet from other technologies or communication networks —is its continued evolution. Its amazing success has only been possible because of its development model built on openness, transparency, decentralization, and its distributed nature. Because the Internet is an open platform, users, network engineers, and businesses of all sizes can innovate both with regard to the platform itself—the Internet–and in how that platform is used… The fact that the Internet remains open to ongoing evolution in its development, operation, management, and use means that the opportunities for context-shifting innovation and creativity still abound today. Innovators are not locked into a centrally predetermined future. Instead, they have the freedom to create multiple possibilities, with success or failure dependent upon whether they are taken up by users.”

The OpenInternet.gov Blog will feature live posts from today’s Innovation, Investment, and the Open Internet Workshop, scheduled to begin at 4:30 pm ET from MIT Media Lab, Bartos Theater (Agenda and Bios). You can watch the event live at http://reboot.fcc.gov/live/. You can also follow the workshop on Twitter using #OiDiscuss.

The Internet has become an unprecedented, global platform for commerce, unleashing the power of American innovators and driving economic growth and job creation. This workshop will examine how the Internet’s openness affects the ability of network operators, Internet content and application providers, and other Internet technology developers to innovate and to drive investment, job creation, and economic growth throughout the Internet ecosystem. These issues will be explored from the diverse perspectives of innovators and entrepreneurs, investors, network operators and equipment vendors, and experts in Internet innovation and investment.

The following will be summaries and paraphrases of the panels, presentations and discussions from the workshop (with apologies to the presenters if I mangle their message).