What I wanted to do with Bellhead/Nethead was focus attention on a hard question:  who should be in charge of the internet?  I emerged with the sense that it is time to stop being incrementalist about this question.  It's apparent that the FCC's current statute doesn't fit internet services well.  On the other hand, it's not clear to me that we should rewrite the act to fit the internet better.  And tweaking around the edges of the act just isn't working (and won't work).

We need to, as a country, take a firm stand and reaffirm the message of 47 USC 230:  no special-purpose meddling.  Of course fraud statutes and antitrust laws and copyright laws will continue to affect online activities -- as they should.  But if we enact special internet-service "social policies," every other country will too.  And if we act as if the internet is a US thing, every other country will act as if it's their thing.  Maybe we want access for the disabled -- another country will want censorship or privacy or who knows what.  Extraterritorial assertions of jurisdiction will become more and more aggressive.  We need to take the lead and restrain ourselves.  Otherwise, we face online mutual assured destruction.

It's tempting for telecom regulators all over the world to say "we should be in charge."  That's what the WSIS process is about.  We should resist all the incremental steps towards that end -- and that means resisting adoption of the agenda set forth in the IP-enabled services rulemaking (the subject of the conference).  We need to start from a blank page, and neither import old regulatory structures into the internet nor set up new internet-specific national (or global) regimes.

Andy Pincus gets it.  If you weren't there, you missed an extraordinary gem of a speech.  I'm very grateful for David Weinberger's stellar blogging -- David captured the essence of Andy's talk:

The Internet is global. We had been successful on pushing our "don't regulate the Internet" line, but with the World Summit on Information Services there's a serious push for world regulation of the Net. Countries have inconsistent demands. To defeat pro-regulation forces, we hvae to take a decentralized approach. We have to defeat them all, not patch them up. The ITU will have exactly the opposite approach.

That's right.  The ITU sees things from the bellhead, centralized control perspective.  The staff at the FCC are smart and kind people, but the small steps set forth in the IP-enabled services proceeding are part of the global trend that ITU is encouraging.  Let's think twice, breathe deeply, and remember that we understand the internet.  We -- the users of the internet -- are in charge.

I'm so grateful to so many people re yesterday's conference.  I thought I'd stop gushing about this as today wore on, but it's 9:30pm and I still feel grateful.  I'm so glad that Michael Herz and the Floersheimer Center wanted to hold the conference at Cardozo.  I'm so grateful that the university was willing to install an open wireless connection just for this conference.  I'm abashed by Weinberger's blogging.  I'm awed by Dan Gillmor's effective, thoughtful approach to online life.  I'm amazed that so many terrific people came -- and sorry that they all got rained on so hard when they left.  Sorry you had to rent a car to get home to DC.  I was so happy to see my cousin Benjamin -- and completely surprised. 

But I think the conference was more than a good time (at least, I hope it was).  It was designed to be a call to arms; a virtual trumpet.  There's a lot to pay attention to -- Induce negotiations, ICANN hearings, summits of various kinds -- but the big picture is the real news right now.  Big government wants to be in charge of the internet, and we have the opportunity to resist on a national level.

If you like music and you like inventors, go see Moog. Bob Moog sees his Moog synthesizers as part of the cosmic order of things -- he never planned that they would come into being, he thought that they would be for experimental music fans who wanted to create new kinds of noises, and he never thought they would be used for tonal, keyboard-driven sounds. One chance meeting led to another chance meeting, which led to a chance demo, and we ended up with Switched On Bach.

  The Moog is an analog instrument.  Bob Moog had no expectation that it would always work, and he says he was often very nervous at concerts.  He talks about being able to feel the sounds by looking at the circuit board, and he's such a gentle, entrancing guy that you have to believe him.  You can tune the circuits and fiddle with the sounds, live.

People hated this analog synthesizer when it first came out.  One interviewer stared at Moog and said, "Don't you feel guilty for what you've done?"

Moog changed the face of music.  He started as a theremin guy, and one thing just led to another.  Sometimes you just can't plan ahead of time.

There's a new Induce draft floating around as of late this afternoon.  It's not likeable; although it more carefully focuses on the showing of intent through conscious and deliberate acts, it still creates a risk of liability for manufacturers or service providers who make things available that permit public availability (eg, providers of peer to peer systems, including IM, and general purpose computers) of copyrighted works.  Initial impressions: 

1.  Great initial breadth: whoever intentionally induces infringement shall be liable as an infringer.

2.  Intentional inducement does require "conscious and deliberate affirmative acts that a reasonable person" would expect would result in widespread infringement.  That's better than actions just being "a cause" of infringement.

3.  Knowledge of infringement (even actual knowledge) is not enough to add up to intent.

4.  Private, noncommercial, non-public-distribution by endusers is not enough to justify a lawsuit under this act --  this may mean a motion to dismiss would work if the only evidence of infringement was this kind of end-user behavior.

5.  Advertising, financing etc. products and services is not the same as intentional inducement -- at least by itself.  These actions plus more could be intentional inducement.

6.  Injunctive relief should be limited to stopping infringing uses -- which means that judges will be in the business of blessing design deals insisted on by plaintiffs.

7.  Pattern or practice of inducement (shades of RICO!) may lead to statutory damages award -- amounts could be huge.

8.  Sony not affected by this act -- translation:  you won't need to sue for secondary copyright liability any more, because the broad language of this act will give you a cause of action for direct infringement liability.

We'll see what happens next.

I was at a joyous pre-IFP event tonight (tomorrow's seminar is all about just how bad copyright problems are for independent film-makers).  It was a gathering of the loyal opposition:  we're all for copyright, don't get us wrong, but we're not so sure that the trend lines are going in the right direction.  We had a fine time.

Perhaps we should have gone collectively (if only virtually) to the content session David Weinberger went to earlier this week.  Froomkin snips from Weinberger's post:

they’re going to win. They own Congress and neither Congress nor the entertainment cartel sees any reason to compromise. Their Lakoffian frame tells them that they’re stopping theft, end of story. So they are going to kill the Internet and they don’t even know it.

If these two groups ever met -- other than at a polite Hill assemblage to discuss the future of the Induce act -- could they have a civilized conversation?  Or is discourse impossible?

Last night I saw a presentation by Pam Gray and Xenogene Gray about their eGanges [pdf] system for depicting legal structures.  Their maps of law use soccer ball icons to show you when you're supposed to go down to a lower level of logic -- say, to see the elements necessary to make a certain proposition true.  Take a look at the pictures.

The eGanges system has got to be only a start.  Laws should be pictures as well as texts accessible through Findlaw.  We should have law museums, so that we can visit old bad laws.  We should have interactive subway maps of law.

You should be able to interact in whatever way you want to with a legal picture -- seeing where your facts fit, swimming in the river of the law, following the train -- whatever metaphor you'd like to use.  The text of laws will remain important; indeed, there could be competing "translations" of law into pictures.  You might want to use a particular picture of a law in your closing argument.  Perhaps that picture will be too prejudicial and your opposing counsel will object! 

It was a fine, creative evening, and it made me want to go back to "Property: The Video Game."  But there just aren't enough hours in the day.

Lessig's debate with David Boies is sold out tonight, and I've spent enough time on Bellhead/Nethead today (please do come -- David Weinberger is the designated blogger, so you can influence what he blogs by saying something particulary eloquent at the mike), so I thought I'd take the opportunity to rise above it all and point out just how important words are to policy.

Yes, words.  You may only care about clickable statutes and pictures of law.  But policy is driven by phrases.

The Broadcast Flag. The broadcast flag is beautifully and effectively named, because it is neither about broadcast nor limited to the waving of a patriotic "flag." Indeed, those who learn about the broadcast flag scheme quickly forget that it is focused on protecting digital television broadcasts and speak generally about the protection of digital content. And the "flag" is, in a sense, the least important part of the entire scheme.  All it does is signal "the following content should be protected."  The heavy lifting -- the encryption and locking-down of the content -- is done by the FCC mandate and by the machines affected by the mandate.

"Unregulation" of IP-Enabled Services. As the internet world has continued to explode, some of the regional Bell operating companies -- heavily regulated by the FCC -- are supporting FCC's call for "social polices" (not economic regulation) to be applied to IP-enabled services. Chairman Powell, in a separate statement accompanying the IP-enabled services NPRM, said "rules designed to ensure law enforcement access, universal service, disability access and emergency 911 service can and should be preserved in the new architecture." 

When Pulver.com filed a petition for a declaratory rulemaking with the FCC, asking that its Free World Dialup (FWD) service (which is essentially an instant messaging service with voice capabilities that does not connect to the traditional telephone system) be declared not to be a "telecommunications service," the FCC responded that FWD was an "unregulated information service subject to FCC's jurisdiction."  The Commission included some ringing language in its Memorandum and Order, saying: "This action is designed to bring a measure of regulatory stability to the marketplace and therefore remove barriers to investment and deployment of Internet applications and services." Chairman Powell's accompanying statement read: "Our ruling formalizes the Commission's policy of 'non-regulation' of the Internet and, in so doing, preserves the Internet as a free and open platform for innovation." 

Some media outlets read this "unregulation" and "non-regulation" language to mean that internet applications would remain unregulated by the FCC.

But "unregulation" does not mean "no regulation." Social policies," including design mandates under CALEA for a subset of IP-enabled services, are envisioned by the FCC to be part of "unregulation."

Other examples? send word.

The bill formerly known as Induce is steaming through Senate processes, and is now scheduled for a markup next Tuesday the 21st (see Ernie Miller and EFF for details) despite thousands of calls from concerned people. 

I spent a couple of hours talking yesterday to a small group about the broadcast flag.  The group came up with a series of hypotheticals -- what if you're a TV manufacturer who refuses to build to the broadcast flag standard?  Are you liable under the DMCA as well as the flag rules?  (Because the broadcast content was never encrypted -- you received it in the clear -- you're not circumventing controls over it.)  What if you're a software developer who sends around freeware that will allow users to get around the flag?  Again, is this a violation of the DMCA?  What if you give advice about how to avoid the flag? Liable, and if so how?  If the suit challenging the flag is successful and the flag regime is declared beyond the FCC's jurisdiction, it's not clear that the DMCA will cover these situations.

What's useful about the bill-formerly-known-as-induce to the content industry is that it provides cover if the flag fails.  It would fill in secondary liability holes left by the DMCA.  The text of the bill is a moving target, but its overall goal is clear:  leave no liability unaddressed.

I am a fuddy-duddy.  This is obvious for many many reasons, but also is demonstrated by the fact that (1) I went to play string quartets tonight and (2) I was twenty minutes early.  I have been walking around the block to avoid being too early since I was six years old.

But being early sometimes has its advantages.  Tonight, I spent my twenty minutes with the Lehman Brothers building at 50th and 7th.

This is a building of flat TV screen displays.  Sometimes ghostly images of employees (there's a loop) walk across the front of the building, but most of the loop is nature scenes with LEHMAN BROTHERS written across them, moving constantly.

Or bridges.  Or mountains.  (These images are from www.lightningfield.com.)

I'm going to take a stand here that demonstrates once again some intrinsic fuddy-duddiness:  I want buildings to be buildings.  This building was trying to give me a rich media, quasi-online experience.  It didn't allow me to imagine life inside the building -- instead, it screamed its brand in constantly moving images that grabbed my attention.  Granted, the building also told me the time -- there's a huge time readout area on the top panels -- and I like that in a building.  But otherwise I have no impression of the building, just a memory of lights so bright that they lit up the entire block.  (My host later told me that Times Square has a "minimum signage" rule to keep things bright and kinetic.)

I like buildings that don't pretend to be web sites.  I don't mind web sites that pretend to be buildings.  I'm a big fan of fully-immersive virtual worlds.  But you know what you're getting and you've chosen to go there.  You're not spending twenty minutes with a blinking building, wondering when the loop is going to start again. 

We ended up playing one of the big Brahms sextets (there are only two, and if you are a real fuddy-duddy you already know what keys they are in), and that was satisfying.  And I still use email.  People under 30 apparently don't do that any more.

If the end to end argument suggests that no lower layer should discriminate against the ones above it, and the First Amendment suggests that no state actor should discriminate against the speech of an individual (except when that speech is illegal), then isn't there an argument that the two are on a continuum?

After all, "speech" at the lowest levels of the protocol stack is pure action -- "send this along." In the US we feel that speech should be protected from state intervention.  When we're legislating about communications networks, should we try to enshrine the end-to-end argument in statutory texts?  We could see ourselves as extending the idea behind the First Amendment to protect fundamental communications from being interrupted by anyone, including but not limited to state actors. 

Or would it be futile to legislate end-to-end?  We might end up drafting tech mandates that limit companies in what services they can offer -- not a very attractive option.

Two big events in US cyberlaw history affecting intermediaries this week: first, a very lively and important decision [large pdf] finding that a Pennsylvania law mandating that ISPs block child porn sites is violative of the First Amendment because implementation of the law leads inevitably to massive overblocking. 

ISPs were blocking IP addresses (shared, sometimes, by tens of thousands of sites) and DNS addresses (shared, obviously, by all the resources at the third level and above) given them informally by law enforcement.  ISPs were doing this blocking in this way because it was cost-effective -- URL blocking (which would have been more targeted) was too expensive and difficult.  And they were doing this blocking nationally, because these ISPs couldn't implement blocking on a Pennsylvania-customer-specific basis.  All of this led to blocking innocent, legal speech. 

CDT and the ACLU challenged the constitutionality of the law mandating that ISPs block child porn sites identified by law enforcement.  These brave organizations were in the tough position of fighting a law addressing the worst human behavior.  Because of how that law was being implemented, it was having a negative effect that far outweighed its benefits. 

As the court says on p.88 of its decision, "More than 1,190,000 innocent web sites were blocked in an effort to block less than 400 child pornography web sites, and there is no evidence that the government made an effort to avoid this impact on protected expression."

A key section of the opinion, on p.91, states that "the Act allows for an unconstitutional prior restraint because it prevents future content from being displayed at a URL based on the fact that the URL contained illegal content in the past." If the Pennsylvania AG requested that a URL be blocked, and ISPs cooperated (and they always cooperate, because child pornography is absolutely abhorrent and illegal), the URL would never be "seen" again. The AG would never reconsider, and the ISPs would never lift the block.

URLs are places, like movie theaters. An illegal theater may become an art gallery -- and it should be allowed to reopen.

Another key portion of the decision finds that the burdens imposed by the Pennsylvania law are violative of the dormant commerce clause (at 99): "Defendant argues that the Act only burdens child pornography, which is not a legitimate form of commerce. To the contrary, the evidence demonstrates that implementation of the Act has impacted a number of entities involved in the commerce of the Internet – ISPs, web publishers, and users of the Internet." Laws directed at illegal activity unduly burden commerce when they have spillover effects on innocent players. Now all we need is a global dormant commerce clause.

This is a lively opinion.  It demonstrates a willingness to understand how the internet works, what ISPs do, and how facially appropriate legal actions may unreasonably disrupt the evolutionary flow of information and commerce.

The second big intermediary event this week was the dissemination of the Copyright Office's draft copyright bill.  This bill (not yet formally proposed) has potentially deadening effects.  Read for all it is worth, it would make copyright infringers out of anyone providing a service (like ISPs) when the service is "a cause" of individuals "engaging in infringing public dissemination of copyrighted works."

Unlike the Pennsylvania decision, which recognizes that illegal behavior by end-users (like posting illegal content) should not necessarily lead to liability for intermediaries where the remedy would have unreasonable spillover effects, the Copyright Office draft would hold all intermediaries and manufacturers liable when end-users infringe copyrights -- no matter what effect this liability has on innovation.  The draft attempts to temper this liability by limiting it to "public dissemination" of copyrighted works (how many people is a "public"?) and by saying that only intermediaries and manufacturers who make money on infringement (how much money?) will be on the hook.

The Copyright Office bill may be a deadly invitation to litigation over the role of technology in our society.  And it clearly doesn't recognize the need to focus on illegal behavior; it outlaws "products" and "services," even if they can be used for noninfringing purposes.  

I'm not saying that the bill, if passed, would kill off innovation.  But it would have substantial unintended consequences that would not be friendly to the continued evolution of technology.  

No one likes to talk about death.  How do we kill off industries who are trying to guarantee their continued survival through legislation?  It's a wholly logical move for them -- their strength lies on the Hill.  But it can't be good for the rest of the ecosystem. 

As we head into more stormy discussions about the future of internet governance, let's keep in mind the differences between hierarchies (like governments) and networks (like the internet).

In a hierarchy, "subsidiarity" is solved by having higher levels be authoritative over the ones below.  In a network, nodes may not be "equal" (networks are usually scale free and subject to power laws), but no node decides for another. 

Hierarchies provide certainty and finality, both desirable things.  Networks make dynamic decisions in a constant ebb and flow of connections, each node vying for the attention of the others.

Hierarchies can last for a long time or can crumble when attacked at a sensitive place.  Networks can heal themselves by routing around broken connections.

Networks are "better" -- in terms of their potential for the creation of complexity -- than hierarchies.  Because of the competition for attention and deference that happens on networks, interesting variations and combinations can happen. In a hierarchy, variation depends on the creativity of the highest levels.  Biological developments -- evolution -- happen in networks.

And everything is biological in the universe.

There is nothing more inspiring and humbling than the study of complexity.  Everything is revealed to be biological, variations competing contextually and endlessly to create an order to which all things tend, but an order that trembles on the edge of chaos.  Great stuff.

The problem is that humans -- and particularly humans selling products and services to the rest of us -- use "complexity" (the word) as an imprecation.  This has its effects, and I'm worried that too much dependence on the metaphors of biology and complexity will backfire. Or, on the other hand, we could work really hard at capturing the joy and benefits of "complexity" (the word) for ourselves.  But we have an uphill fight in front of us.

Let me illustrate.  I recently listened to a debate among Craig Mundie, Halsey Minor, and Larry Babbio at the August 2004 PFF summit on the Future of the Internet.  In answer to a question about where opportunities lay in serving consumers vs. businesses in the future, Halsey M. talked about the fact that 90% of online traffic is now coming from consumers.  He said that there is a huge opportunity to make the internet more productive for businesses, and that we'll be retooling businesses to take advantage of the internet's possibilities.

Then he began to analogize [broadly paraphrasing]:

Just as the utility grid allowed centralization of complexity -- and the industrial revolution allowed people not to have to worry about producing power any more -- the internet allows businesses to consume centralized complex services from others.  Google centralizes the complexity of search.  This means that companies can stick to their core competencies.

Craig Mundie agreed with him, saying that the complexities of infrastructure and identity mechanisms would be overcome by a few very large companies, and adding that "Without this scale [of services being provided by large companies], businesses can't change and grow."

These were eminently reasonable comments.  But they point to the "complexities" of identity management (a subject focused on particularly by Mundie) being handled by a few very big players, and, generally, to "complexities" created by the internet being smoothed for the consumption of businesses and consumers.

Complexity = bad for business.  That's the lexicographical development.

Complexity involves, by its nature, lots of choices, and a good deal of confusion if you are trying to predict the future.  But if you look at a complex system from the outside, you see emergent order - like the miracle of food delivery happening every day in the city of New York.

If identity management is something that is too "complex" for individual businesses to handle -- much less individuals -- and the perception is that it must be given over to a few large entities in order for businesses to function, we're heading in the wrong direction.  We need to take back the language of complexity -- or perhaps come up with another word.

The Copyright Office redraft [pdf] of S. 2560 got a lot of attention this week.  Essentially, the redraft suggests that if you do a single overt thing (raise an eyebrow? build a microchip?) that could (a) be "reasonably expected" to (i) cause an infringement or (ii) persuade someone to infringe, you're liable for direct infringement yourself. 

"Overt acts" aren't defined, but there are specific things listed that aren't overt acts -- like providing phone service to a "distributor of dissemination technology." Thanks.

The last time I wrote about the Induce Act, I suggested that we shouldn't jump up and down and attack it -- that Congress should hold a hearing and get people to explain why the bill would or wouldn't work.  That hearing happened in late July.  Most of the people testifying, save for the Copyright Office and the RIAA, said that the bill shouldn't go further.

Now the Copyright Office has circulated a redraft that, if anything, is likely to get people even more worried than they were about the original S.2560.  It doesn't seem like a move towards compromise.  If anything, it signals a hardening of position:  any technology that makes infringement possible (not just KaZaa or Grokster) can be reached under the draft, and the Sony/Betamax rule is dead. 

Sony is dead because the draft says that one possible "overt act" could be distributing a technology "that, when used as intended, automatically causes the user of the technology to infringe copyrighted works without the user making a specific, informed decision, for each copyrighted work at issue, about whether to engage in such infringement."  This boils down to:  if you build a technology that makes it possible to distribute works publicly (broadband access? mp3 players that connect to the internet? PCs?) you're liable.  Even if you don't meet current judicial standards for contributory or vicarious liability.

What's remarkable here is that the content industry feels that Grokster has given it an opening to broaden copyright liability beyond recognition.  Why haven't the comments of the many other businesses and advocates that opposed S.2560 been listened to?  Why are people stuck negotiating a special-purpose bill that doesn't seem to be special-purpose at all -- but, instead, seems to take on anything that might be used by an infringer?  Why is the Copyright Office (clearly not neutral on this subject) holding the pen?

I'm sure there will be many meetings about this draft, and I'm confident that reason will prevail.  I'm sure this redraft isn't the last word.  So I'm not jumping up and down.  I'm just amazed at this sequence of events.

Today is the 35th anniversary of the day the first node in the ARPANET (at UCLA) sent a message to another computer.  Next month will be the anniversary of the day that first node talked to the second node. 

And what was the first message sent from one node to the other?  It wasn't "was hath God wrought."  It was one computer saying LOG and the second receiving LOG and adding IN -- so they'd have LOGIN.

In Leonard Kleinrock's words:

We sent an L; - did you get the L;? YEP

We sent an O; - did you get the O;? YEP

We sent a G; - did you get the G;? CRASH!

That was the first message on the internet.

Mitchell Waldrop's The Dream Machine, in a section titled "ARPA's Woodstock," has Doug Engelbart saying the following on December 9, 1968: 

'The research program that I'm going to describe to you,' he began in that soft, strangely compelling baritone, 'is quickly characterizable by saying, "If, in your office, you as an intellectual worker were supplied with a computer display backed up by a computer that was alive for you all day, and that was instantly responsive to every action you had, how much value could you derive from that?"'

Happy many anniversaries.