A less open, less democratic internet?

I'm flying home to (my new home in) Silicon Valley after participating in a 3 hour workshop held by the FCC yesterday in Boston. The topic was the future of the "open internet" and also whether Net Neutrality rules that have guided the internet for most of its life should extend to the emerging "mobile internet."

I was lucky enough to be one of two start-up entrepreneurs invited to speak at the Forum, and represent mobile application developers in general and the barriers we still face.

I'm trying to keep an open mind to some of the proposals that are being floated but I find them still in my gut to be deeply troubling.

Several companies like Alcatel-Lucent and Camiant proposed somewhat radical proposals for varying “quality of service” on the internet. This is code for a system where web site owners or Application developers might have to pay extra to carriers to get faster data transit or lower latency (response time) than others get. This is distinct from the internet today, which is on a "first come, first served" basis known as "best effort" delivery.

My concern is that the internet has fostered incredible wave of new companies, innovation, job creation, and better products for Americans (and people around the world). It's been a pretty open place. Google, Ebay, YouTube, Facebook, Yahoo, Skype and others are great new companies that began often in dorm rooms or garages. They didn't have a lot of capital to start, and when they got venture backing, the costs of building an internet service was relatively low compared to the potential audience.

Under conditions where a big company like Google (now a goliath) or NBC gets to buy 1st class transit, what happens to start-ups and small businesses? Their applications will likely load even slower than today (network capacity is fairly zero sum except in the long term), since an "express lane" has been carved for the fat cats. Remember, most jobs in the US are with small business, and all the net job gains lately are from small business. It's risky to the tradition of the internet as an open forum for competition and new ideas to make it harder for new entrants, especially those that need to prove themselves and get to scale before they can fire up a revenue model (think Twitter or YouTube, or in our case Skyfire).

This "quality of service" billing is under serious consideration. Industry working groups and big cable and telephone companies, along with their equipment vendors and software vendors, want the FCC to clarify rules so they can start doing this. What's held them back has been fear of FCC action. Since the 1990s the FCC has in one form or another held internet carriers to rules or guidelines preserving open access, and in 2005 these open access rules were codified by the FCC into "net neutrality" principles, which basically means everyone's bits on the internet get treated equally, with no discrimination, as long as they are lawful (spammers or viruses, not invited). 2005 was not a change; it basically codified how things have been since the 1970s and the birth of the internet.

These new "Quality of Service" proposals seem quite worrisome for competition by start-ups like mine, and would favor big established companies with lots of cash to “buy up to first class.” Since it’s a fixed pie of bandwidth in many ways, it will have the impact of degrading throughput for data in "steerage class" for the rest of us. Now the proponents of QoS do acknowledge that it should not be a totally unregulated caste system. Based on what Alcatel and Camiant proposed at the FCC workshop, there could be rules to keep this class system somewhat fair. For instance, telcos would have to publish pricing and offer it to any buyer, with no volume discounts or favoritism, so little companies can buy in to first class at the same prices as say Google/YouTube or NBC… but I don’t know that we’d have the money… or at least, start-ups would have to raise more money if they wanted to do video or low latency applications, and that would make it harder for the next YouTube to emerge.

I do want to express sympathy for the private companies like cable and telco players that have invested billions to give us the wonders of the internet, and interconnect to each other to make the information superhighway possible. They have created a wonderful public good. And like any public good, there is a risk of the "tragedy of the commons," as in an open ocean fishery that gets overfished, a common pasture that gets over-grazed, a free highway that gets congested, or now wireless and fixed bandwidth, especially in the local loop. Skyfire has long said that the growth in wireless data consumption is exponential and it's impossible for spectrum and networks to expand fast enough to keep up. Something does have to change. But let's not throw out the baby with the bathwater.

Rather than a radical solution, how about trying a more incremental solution that can probably solve the problem, but stay closer to the openness that has fueled competition on the internet since the beginning? My modest proposal is this: Nondiscrimination should remain the hallmark of network management. Allow ISPs and access providers to offer CONSUMERS the option to buy tiers and flavors of service, by the month or by the minute. Prices can vary for peak and off-peak. It's usage-based pricing. If the user wants really high quality Skype video conferencing, or fast Hulu video, or a "gaming mode" which gives small packets fast transit (low latency), then they can buy that on the spot. Browsers and applications can tie into open APIs to resell this upgraded service to consumers. I do NOT think that publishers or applications should be allowed to buy this higher tier for consumers, as that will distort the level playing field.

Several benefits:
(1) In the industry, it's well known that a few "data hogs" among users really tax the network at the expense of the rest of us. Some of them are doing unattractive things (like uploading hundreds of pirated videos, adult content, spam or viruses). Some are running small businesses or organizations from consumer accounts. Some are music or entertainment fanatics who download huge HD files in high quantity. I'll set aside as a separate topic that the open internet and net neutrality have always been about "lawful" content and so I don't have much sympathy for full-time pirates. But just saying to intense users, "you have to pay up for a "pro" edition of access level," would probably restore market incentives, fund increased bandwidth, and cut down on abuse. And the 98% of users who aren't data hogs will be better off in service and pricing.

I heard from a Cisco friend that a few hundred users in Germany at one point were over 50% of the network data! Those folks should pay for Pro access. Unlimited one-size-fits-all consumer rate plans are unsustainable and ripe for abuse. That's true of both mobile and fixed-line internet both.

(2) The "user pays for faster service" model is as non-discriminatory and close to the open internet as possible, while starting to create tiering. If sold by the hour or minute, it probably won't be that expensive for a casual user. And it won't bias the system in favor of big established companies with lots of cash to buy an "express lane" all the time for their products. And it can have flavors-- some may want "big bandwidth" mode, others "low latency" mode for gamers.

Marcus Weldon, the CTO of Alcatel who spoke at the forum, is a very smart professional and while I may not fully agree with everything he proposed, I completely agree that bandwidth usage is exploding and some way to balance supply and demand is needed.

You can find his presentation and mine on this FCC page:

http://www.openinternet.gov/workshops/innovation-investment-and-the-open-internet.html

I admit that I have more to learn about Quality of Service proposals and there are certainly better and worse ways to implement them. But from the many people who came up to me after my presentation to thank me, I feel I'm right to be wary of QoS. At the very least, it deserves a lot of scrutiny and debate to shape it. What do you think? Comments welcome.

I'm glad I made the trip, and at least made my case on behalf of entreprenuers and VC investors. Kudos as well to Barbara and Shane, two professors at the workshop who made much more scholarly and erudite presentations that concorded fully with my view. Thanks for your good work.

I'll try to post soon about bringing Net Neutrality to extend to the mobile web, which is still under debate before the FCC. Right now, there really are no rules on wireless. And I want to write a bit about my impressions from CES last week. Off to a start on my blog.

Thanks for reading.