Category Archives: networks

SXSW spam? Blame the network gravity well.

Adman Bob Knorpp has been poking fun at the SXSW “panel picker” — a crowd-filtering technique in which democratic votes are supposed to help select the best panels for the Austin digital-media conference, but which instead has devolved into “please-vote-for-me” indignities. Sure, only 30% of the decision comes from voters like you, but how can anyone realistically judge 2,401 panel contenders?

SXSW is proof that social networks can’t duplicate democracy — because, unlike voters, not all nodes in a human network are created equal. (Voting networks don’t work well, either, which is why the U.S. founders set up a Congress with representative experts to filter decisions away from the sometimes-hysterical masses.) Edward Boches, creative chief of Mullen, has 12,763 followers on Twitter and thus gets noticed when he complains about Marriott. Blogger Chris Brogan has 150,485 followers. Both are likely to get voted in if they float a panel, however brilliant or stale the concept. If the SXSW goal is to build a meritocracy of ideas, and no one has the patience to judge 2,401 individual entries, then what remains is a popularity contest.

The gravity well of networks.

We call this network lock-in — a form of groupthink that emerges when networks reinforce their current structures, similar to the stardust coalescing in gravity wells to form a new sun and planets in orbit. Humans are drawn to ideas, and they chase other nodes who espouse their own ideals most fervently. This like-drawn-to-like dynamic explains the rise of extreme news (Fox, MSNBC), horror-movie porn, punk rock, shouting politicians, and uber-bloggers. Fragmenting media blows self-reinforcing communication bubbles, where you, if extremely conservative, can find videos and commentators explaining why Obama is a socialist; or, if liberal, find an equal number of talking heads explaining how Obama is saving the economy. We are lured by gravity; tidal forces pull us to nodes that take power from our joining masses; the extremes of commonality rule the day.

Unfortunately, for edgy forums such as SXSW, this means a concept on the fringe of public consciousness — stardust far afield — may be ignored in favor of the topics already talked the most about. In the interactive series categories this year, we see 136 proposed panels on “social networking,” 77 on “advertising,” yet only five on tablet computers and none at all about artificial intelligence (they are there, but you have to dig). Tablets are the edge of media; AI may change the world soon. Unfortunately, few at SXSW may talk about that future.

Equal rights don’t exist on the internet, do they?


You’d think our government would take a stand on internet access, but guess what? They punt. Yesterday the FCC launched a fuzzy new blog that talks out of both sides of its mouth. It’s all about the issue of network neutrality, which seems simple at first: the internet should be free and open — no variable pricing, no limits on access, no one blocking your content. Sounds good, right?

Before we explain the FCC’s double standard, let’s dig deeper to see the flip side of unshackled internet use — the painfully slow email downloads in your home office thanks to the teen next door downloading fat movie files, the fact that the United States trails Sweden and Korea in bandwidth speeds. If you’ve ever waited for downloads, you realize the web has costs. The words you’re reading now, on this blog, aren’t free at all, but rather the lucky result of over-building during the 1990s internet bubble. You can watch videos at Hulu today because silly companies got silly stock options back in 1997, and investors jacked up the network infrastructure. And guess what? Your prepaid pipes are getting full.

So the internet issue has two sides — keep all access equal, or allow providers more control to raise money for the strapped global network. As with any scarce resource, the network neutrality concept has fierce advocates (Google, Yahoo, Microsoft) and opponents (your cable company), and no wonder — the companies who profit by sending stuff free over pipes want it kept that way, and the firms that pay for those pipes want to be able to charge more to keep them working. In perhaps the strongest argument against network neutrality, telecoms argue that if they could charge heavy web users more for access, they’d be able to fund future advances in internet technology. That is no more unfair that Apple charging $599 for its first iPhone to fund the production build-out that now allows you, two years later, to pick one up for 99 bucks. Early adopters, like heavy internet users, had the most need, and they were willing to pay more — a wealth transfer that ended up supporting the rest of the population. So why not?

Who’s right? The FCC says everybody.

So back to the Federal Communications Commission. The new FCC OpenInternet.gov blog sounds, in name and tone, like it’s promoting network neutrality. Chairman Julius Genachowski speaks with phrases such as “the fifth principle is one of non-discrimination — stating that broadband providers cannot discriminate against particular internet content or applications.” But wait — Genachowski quickly adds caveats. “This principle will not prevent broadband providers from reasonably managing their networks. During periods of network congestion, for example, it may be appropriate for providers to ensure that very heavy users do not crowd out everyone else.” Um, what?

Genachowski concludes the FCC will make decisions “on a case-by-case basis.” So … cable companies and telecoms should not discriminate against any internet users … unless they need to. They shouldn’t charge more for usage … unless required to reasonably manage their network. The FCC will enforce this … with ad hoc decisions.

If you’re confused, so are we.

Perhaps it’s too much to ask that a government bureaucracy take a clear stand on internet access in an age where suggesting we expand health coverage gets seniors, now on government-funded Medicare, screaming about government socialism. The only clarity we found on the FCC’s site was a link to a Digg-style ranking of user suggestions. Here, the public’s wishes rise to the top. Bring U.S. broadband pricing in line with the rest of the world. Promote telecommuting. End unreal claims of internet speeds. Heck, catch up with Korea’s pipe speeds, where people stream TV shows on cell phones.

The public wants bandwidth, and they want it fast. The FCC says it’s listening and it wants “freedom”. But as with any public good, we suspect eventually someone will have to pay for it.

Metcalfe, say it ain’t so


This just in. BT researcher Bob Briscoe, math prof Andrew Odlyzko and programmer guy Benjamin Tilly say Metcalfe’s law is wrong. You know: the basic idea that the more people use something, the more exponentially valuable it becomes. One plane going to one place isn’t worth much; but knowing that plane can connect you to almost anywhere in the world is worth a lot. Robert Metcalfe, inventor of the Ethernet, dreamed it up and social media hyperbolists now tout it as gospel.

The value of a network is proportional to the square of the number of its users.

But Briscoe, Odlyzko and Tilly suggest that different types of networks have differing increases in value. Networks that can form groups — email lists, or Twitter, or MySpace — can explode in value as more join. Others, such as a network of users signing up for a cable system, grow more linearly.

To put this in plain English, if networks really did jump in value at an accelerated rate for every person or node you add, eventually adding one additional person to, say, Facebook would add value equal to the entire global economy. Briscoe and team note common sense tells you that can’t possible be true. Or, in a business sense, if adding to a network increased value exponentially, then every communication group in the world should merge with another one — since the combined larger network would be valued infinitely more than the two standing alone.

Look, if this gives you a headache, we’re sorry. Just assume the following: Not all social media bubbles rise forever, and not all 2n joined systems are worth (2n)2 or 4n2. Microsoft and Yahoo, are you listening?

(Illustration by Serge Bloch. Inspiration via Fake Steve.)