Kickstarter provides a way for creators to route around the strictures of traditional publishing industries for a variety of reasons. We've seen Double Fine break the bank on Kickstarter, after getting a lot of blank stares from publishers who didn't think there was a market for point-and-click adventure games. We've seen top Hollywood talent fund a film on Kickstarter to ensure no studio execs would be meddling with their movie. Now, in what I think is a first, we've got a parody author turning to Kickstarter after his publisher dropped the project, citing fear of a legal backlash.

Michael Gerber already had some published parodies under his belt (most recently the Barry Trotter series, from Gollancz, which also publishes spoofs like Bored of the Rings) when he was approached by a major publisher for his take on TV drama Downton Abbey. The writing was complete and the contracts were almost signed—but as he puts it:

And then somebody in New York started… thinking.

“What if [Downton Abbey creator] Julian Fellowes gets mad?”
“What if they hire another company to publish the official Downton Abbey Calendar/Tea Cozy instead of us?”
“What if I make the wrong decision and get fired?”


So even though Downturn Abbey is the most fun I’ve ever had writing a spoof, Mr. Big Publisher changed his mind and offered me a couple bucks to shred the manuscript.

The hell with that! I like this book too much to kill it.

So, rather than accept the kill fee, Gerber is trying to self-publish the book with crowdsourced funds—and with over 50% funding only a few days into the campaign, it looks likely that he'll hit his (relatively modest) goal of $3,500. It's interesting to see a publisher's legal panic serving as the prime motivation for a Kickstarter project. In the "Risks And Challenges" section that Kickstarter now requires on all projects, Gerber directly addresses the legal question, and takes an admirably principled stance:

Now, some of you might be thinking, "Won't he get sued?" I've been doing this professionally since 1991, and so far, so good. Parody — particularly print parody — has a very robust precedent, and I've specifically written Downturn Abbey to fall within it.

...

A right only exists if you exercise it, and one of the biggest reasons I'm crowdfunding this project instead of taking the kill fee is to assert the right to critique our shared culture via parody. Media corporations will extend their rights relentlessly, so if you like parody — in any medium — it's important to support it in print, because this is where it is most vulnerable. Nobody's suing Jon Stewart, but if parody gets rolled back, that's the logical next step. Downturn Abbey is a harmless bit of fun for fans of the show and the era, but if it can be suppressed out of corporate risk-aversion, sooner or later that attitude will have a chilling effect on the rest of comedy.

Unfortunately, due to the nature of copyright lawsuits and fair use defenses, it would cost a lot more than $3,500 to defend his stance in court. Hopefully, though, it won't ever come to that—the creators of (and anyone else with a copyright stake in) the show would hopefully realize that there's little point (and less honor) in trying to squash out parodies of this nature. However, Gerber's move also highlights the associated problem: even when something is as well-protected as parody, copyright law can have a huge chilling effect by making publishers nervous. Kickstarter and other tools for creators are, at least, helping to combat this—for a publisher, standing behind a parody is just another risk factor in the spreadsheet; for Gerber, it's a matter of standing up for his rights.

Most of us have been taught to understand math-related topics in a linear way, but that might not be the way our brains are hard-wired. Kids actually tend to have an innate number scale that is logarithmic, so even though they know how to count to ten (or even twenty), they'll actually think more along the lines of one, many, lots of many's, and then OMG so many that's like infinity. Here are just a few links on logarithmic thinking to ponder.

• Ask second graders to map out a few numbers on a blank line from 1 to 1,000 -- and you'll see that 7yo kids tend to space out the numbers in a logarithmic pattern. Fourth graders won't make the same number line map as these second graders, and it's apparently not too hard to un-learn logarithmic scales since many adults don't remember what a log scale graph is. [url]
• RadioLab has a nice podcast about how kids learn their numbers, switching from logarithmic thinking to more linear thinking. Young kids and even infants notice large changes (such as when quantities double), but they're not so good at exact amounts and small differences. [url]
• It's not just little kids. Some cultures think in logarithmic scales -- such as an Amazonian indigene group, the Mundurucu. Actually, logarithmic thinking may lie dormant in all of us, whenever we estimate quantities or deal with extremely large numbers. [url]
• What number is halfway between 1 and 9? Most formally educated adults say 5, but another answer (from kids) is 3. [url]

If you'd like to read more awesome and interesting stuff, check out this unrelated (but not entirely random!) Techdirt post.

Tech is all about disruption. When a problem presents itself, someone will find a way to route around it. Red light cameras, the controversial jackpot generators favored by certain law enforcement agencies, are the problem. The solution?

Jonathan Dandrow has developed noPhoto, which renders the pix snapped by those revenue-generating robo-cams useless. The technology behind noPhoto is fairly simple. At the top of the gadget, which doubles as a license plate frame, there’s an optical flash trigger that detects the flash of the traffic-light camera. That trigger sets off one or both xenon flashes in the sides of the noPhoto, so when the traffic-light camera opens its shutter, there’s too much light and the picture of your license plate is overexposed. Big Brother can’t read your plate.

Some will argue that technology like this will only be used by scofflaws wishing to run red lights. But as the inventor points out, there's plenty of shady activity on the other side of the camera, not the least of which is some more erosion of civil liberties.

“I just had a lot of reservations about the cameras,” Dandrow says. “They are trying to circumvent the constitution.”

As Gary Biller, the president of the National Motorists Association, recently wrote in U.S. News and World Report, traffic-light cameras violate “several key tenets of a citizen’s due process rights,” because there is “no certifiable witness to the alleged violation,” and so therefore, “the defendant loses the right to cross-examine his accuser in court.”

This isn't just an overreaction from minor league anarchists. Once the camera has "decided" you're guilty, that's it. Arguing your innocence is next to impossible. Case in point: this woman's ordeal with being falsely accused by a red light camera.

The New Star-Ledger’s Bamboozled column has the story of a woman whose license was suspended by authorities in New Jersey for a red-light ticket, even though the car in the photo is clearly not hers — oh, and she hadn’t even been in the Garden State in more than a year when the violation occurred.

Despite having moved from New Jersey to Colorado 18 months before being ticketed, she still had to jump through a ridiculous number of hoops to clear her record and get her suspended license back. Starting with the city staff in Edison Township calling her a liar and a long game of paperwork tennis between the New Jersey Motor Vehicle Commission and the township, it finally reached the point where the driver realized it would be easier to pay the fees than travel 1800 miles for a court date. Appealing the ticket runs $75 plus fees and a personal appearance. The final breakdown? $200 to restore her license plus $81 for the ticket and $33 in court costs. $314 to clear a ticket issued to the wrong person.

Despite the obvious flaws in the "camera-as-cop" system, there's a very good chance Dandrow's noPhoto will be forced off the market before too long. There's already a law against obscured license plates. But will that hold when the license plate is only obscured when being photographed? The other angle will likely be "concern" about "safety," as if this bit of civil disobedience will lead to rampant red-light running. Will some people ignore red lights if they have this equipped? Sure, but I doubt they'll use it to drive through heavily-trafficked intersections against the lights. This will most likely result in a few more "California stops" at quiet intersections. If the cops don't like this, maybe they should consider the example they set when they flip the light bar on momentarily to blow through traffic signals.

If this disruption hastens the demise of the rightfully-maligned red light camera, so much the better. The cameras, which some municipalities adore for their money-making ability, have been linked to increased accidents and shady law enforcement activity, like issuing tickets signed by a deceased patrolman, lowering the yellow light time to increase the chance of violations, and lying about the length of a red light to hide the fact a bogus ticket had been issued. Fighting back against technology being used badly for lousy ends with a simple, effective solution is the best kind of disruption.

Mike has been advocating an independent invention defense against patent infringement for quite some time now -- that is, in the absence of actual copying, a patent holder should not be able to sue a subsequent innovator. There are plenty of posts going over the rationale for such a defense, and I don't want to rehash those arguments when you can just click on those links above and read the originals.

What I do want to do is address one of the most common concerns with the independent invention defense: feasibility -- in particular, the concern that independence (or dependence) would be too difficult to prove. See, for example, some of the comments on the EFF's push for this:

"Independent arrival" is a vacuous notion, with possible noble intent, but difficult-to-impossible to prove. It would require mind-reading beyond current human abilities.

Or this one:

How on EARTH would a patent holder succeed in court if the defense could just claim ignorance?

The short answer is that we do this already. Patent holders can claim extra damages if they show that an infringement was "willful", and occasionally they succeed in proving this.

The long answer is that our legal system has been dealing with this problem for a long time. Someone shoots and kills someone else. The shooter claims self-defense. Or that it was an accident. Or insanity. You'd think that these defenses would "require mind-reading beyond current human abilities", but as it turns out, judges and juries have done an OK job at sorting it all out. Sometimes people get away with murder, but on the whole, murders have not skyrocketed because of self-defense claims.

One reason is that it's harder to lie in front of a jury than many people think it is, especially when the other side's lawyer can call you out on it.

The other reason is that there's evidence outside of what you tell the jury. If you murder someone, there are witnesses, security cameras, smoking guns, and friends that rat you out. If you're an inventor, you make notes and schematics. You communicate with co-workers and colleagues. You browse the Internet, leaving a browser history and IP log. And if you work for a large corporation, all of your communications and notes are likely logged and archived precisely because of the risk of litigation. That's a lot of evidence that could be used to prove, or disprove, independent invention. Just look at the Apple v. Samsung case, where the jury was able to rely on Samsung's own documents to conclude that Samsung had willfully copied Apple (whether such copying is really a problem is a different question).

Independent Invention and Reproducibility

A corollary of needing proof of dependence is an incentive for patentees to disclose their inventions in a manner that's actually useful to others. The law currently requires patentees to describe enough of their invention to enable someone else to reproduce it, but they certainly don't have any incentive to make it easy to reproduce. Trolls actually have an incentive to obfuscate what a patent covers and spring it on a business only after that business has sunk billions of dollars into developing their own product.

By contrast, if a patent holder can only enforce his patent against parties that actually copied the invention, the patent holder has a strong incentive to disclose in a way that makes copying easier. Perhaps patents would look less like this IBM patent for "[s]ystems and articles of manufacture for managing metadata associated with a data abstraction model abstractly describing data in a database" (ugh) and more like this Ruby on Rails tutorial, which helpfully includes screenshots and code samples. This improved disclosure would make it a lot easier to prove dependence. For example, if your patent includes code samples, an infringing product might contain code suspiciously similar to your own. Or if you included useful links, the infringer's browsing history might reveal frequent visits to those sites. And so on.

Focusing on the actual usefulness of a patent would also address another concern with an independent invention defense -- that trolls would spam businesses with patents to demonstrate that business had notice of the patents, and that businesses would avoid liability by forbidding their employees from reading patents altogether. If a patent plaintiff had to show not merely that the defendant knew about the patent, but that the patent was actually useful to the defendant, the trolls would be out of luck. Usefulness is not part of their strategy. Likewise, if patents were useful, many businesses would take advantage of them. Independent invention, like all invention, takes time and money. So long as it's cheaper to license than to independently invent, (smart) businesses will choose to license.

Dealing with Causation Problems

A related concern with proving independence is that nothing is truly independent. As Mike often points out, innovation isn't about creating something entirely out of the blue but it is about taking existing ideas and improving them. How then do you prove independence when the innovative process is so often about copying? And what if earlier patents did contribute to your invention, but the contribution was infinitesimal?

As it turns out, this is also a familiar problem for the legal system. It's very similar to questions of causation. Traditionally, a defendant was only liable if his actions caused the plaintiff harm -- that is, but for the defendant's actions, the harm would not have occurred. Likewise, you might say a later invention is dependent on a prior invention if the later wouldn't have been invented when it was but for the prior.

The but-for test has some obvious problems. Suppose two campers start fires on opposite ends of a forest at the same time, and both fires spiral out of control and burn down the forest. You can't say, but for the first fire, the forest wouldn't have burned down, because the second fire might have caused the same amount of damage. Or suppose a liquor store clerk sells alcohol to a minor, who is then caught by her morally uptight mother, who is in turn so mortified than she runs outside and is struck by a speeding driver. But for the sale of the alcohol, the mother would not have been hit, but it seems silly to blame the liquor store clerk for that.

The legal system basically addresses this problem by using open-ended language and granting discretion to judges and juries on edge cases. Thus, in the case of the forest fire, a jury could hold both campers responsible because each fire was a "substantial factor" in the burning of the forest. And in the case of the uptight mother, a jury could give the liquor store clerk a pass, because the liquor sale was not the "proximate cause" of the mother's injuries. Alternatively, the legal system may give judges and juries the discretion to recognize differing degrees of fault and adjust damages accordingly.

As with any system that defers to personal judgment, some decisions may seem arbitrary, but for the most part, causation problems have not crippled the legal system. And if such problems were encountered with an independent invention defense, the outcomes would probably be less arbitrary than in tort cases. For instance, there's often no objective basis for saying a driver is 60% at fault in an accident as opposed to 70%. But, in a patent case, records of profits, research expenditures, and the costs of alternatives provide evidence of how much a given patent actually mattered.

Making Patents Useful

Patent advocates might object that an independent invention defense would introduce greater uncertainty into the patent process, and that it would hurt small (independent) inventors more than large corporations. First, an independent invention defense, depending on how it's drafted, could potentially reduce uncertainty in other areas. For instance, how much one invention contributed to another is less subjective than asking whether the invention was "obvious". Second, most legal changes tend to introduce uncertainty, at least initially. And yes, smaller inventors will be affected more simply because they're less able to afford expensive lawyers. But sometimes uncertainty is necessary to achieve reasonable legal outcomes. Adding a self-defense exception to laws against murder introduces uncertainty as to how a case will turn out. But nearly everyone agrees that some right to self-defense should exist, even if the boundaries of that right are unclear.

Moreover, these objections are often less about any issues with the defense itself and more about objecting to any change that makes patents less valuable. Likewise, some Techdirt readers would object solely on the basis that independent invention falls short of simply abolishing all patents (or at least all software patents). All I can say to either position is that the independent invention defense reflects the view that patents are currently too powerful, but that maintaining some level of patent protection makes sense (or is more politically palatable than total abolition). It's in line with Judge Posner's thinking that patents are useful when inventors face high capital expenditures to create, but low barriers to copy -- an independent invention defense limits infringement to actual copying. It also addresses Tim Lee's concern that the boundaries of software patents aren't clearly marked -- an independent invention defense would protect programmers who inadvertently infringe on those patents. And it solves the one-size-fits-all problem with patent lengths -- more obvious patents will last only a short while before they're independently invented, while less obvious ones will last a bit longer.

As with all laws, the devil is in the details. An independent invention defense would have to be drafted carefully. But the basic legal concept isn't new. It's feasible, and it could go a long way towards curbing patent trolls.

What is it with hosting companies who are quick on the trigger to take down entire sites in a kneejerk response to legal threats, going way, way beyond their legal obligations? We recently wrote about hosting firm ServerBeach taking down 1.5 million blogs over a single copyright claim (when to keep their DMCA safe harbors, they only needed to take down the one bit of content highlighted). Now another hosting company, PhoenixNAP, has done something even more ridiculous. In response to a takedown notice (pdf) sent by Atlanta mayor Kasim Reed's lawyers, PhoenixNAP took the entire gossip site LipstickAlley offline without any notice to the site owner.

Reed was arguing that comments on the site were defamatory, though it's telling that his lawyers went straight to the hosting company, rather than the site itself. Even so, PhoenixNAP was under absolutely no legal obligation to comply. Unlike the notice-and-takedown provisions of the DMCA, when it comes to defamation, hosting companies have even greater safe harbors under Section 230 of the CDA, which makes it clear that service providers are not liable for speech made by others. Period. They don't have to take down the content. If Reed wants to sue those who actually made the comments (which are Lipstick Alley users, rather than Lipstick Alley itself), that's his business, but PhoenixNAP is under no obligation to do anything.

Amazingly, even after being informed of this, PhoenixNAP stood by its decision (with the exception of noting that it should have informed Lipstick Alley's owner that it was cutting off the site). As noted by Public Citizen's Paul Levy:

In response to a strong protest, Phoenix NAP acknowledged that its failure to give notice was a mistake in process, but it had no sympathy for Lipstick Alley’s legal rights; PhoenixNAP told me that it takes claims of defamation seriously and, without regard to the merits of the dispute, its customers must “resolve the issue with the complaining party.” Indeed, PhoenixNAP was not at all disappointed to learn that Lipstick Alley felt it could no longer continue as a hosting customer, because, apparently, PhoenixNAP believes that web sites on which users can post comments generate too much trouble. The discretion accorded to hosting services to avoid certain kinds of web sites is the other side of the coin from the important protections that section 230 affords, but consumers should be aware of the limitations before they are induced to sign on as customers. PhoenixNAP might consider doing a better job of explaining its preferences

These days, if you're hosting a website, you want a hosting company that will stand up for your rights, and recognize its own rights as well. This move by PhoenixNAP has made it clear that it doesn't stand up for its own customers, and it should raise alarms for anyone who hosts a website with PhoenixNAP. Apparently the company will shut you off if anyone complains, and it's your problem to deal with it. That's not a hosting company that I would want to work with. Furthermore, as Levy explains, actions like this only encourage more complaints to be sent their way:

You would think that a hosting service like ServerBeach or PhoenixNAP would respond with hostility to complainants who take their demands straight to the hosting service instead of beginning with the web site where supposedly improper content is hosted. When the service responds directly, and especially when it responds by taking down the customer’s entire site, the service not only encourages others to impose on the service by complaining there instead of to the underlying site. The service also risks losing long-term customers who think that they ought to be given a bit more respect.

It is not only hosts of message boards or bloggers who allow comments who should worry about PhoenixNAP’s attitude about mere claims of bad content. With PhoenixNAP playing the role of super censor for any web site it hosts, whenever an unhappy target of criticism takes its complaints straight to the data center, no web site operator can be confident about using its hosting services for sites that discuss public issues or public figures in ways that those who can afford to hire lawyers to send threatening letters may not like.

As for the complaints of defamation against Lipstick Alley itself, the site has made it clear to Reed's lawyers that it is also protected by Section 230 (pdf) and that the site "does not negotiate with bullies", such as those who send questionable takedown notices to web hosting companies rather than going to the appropriate parties.

Earlier this week, we talked about how publishing giant Random House had very explicitly stated that when libraries buy their ebooks, the libraries "own" those ebooks, rather than license them. They left no doubt about it. Skip Dye, Random House's VP of library & academic marketing and sales was explict: "when libraries buy their RH, Inc. ebooks from authorized library wholesalers, it is our position that they own them... this purchase constitutes ownership of the book by the library. It is not a license."

This raised some questions, such was whether this was true of everyone else who purchased Random House ebooks. Peter Brantley asked a bunch of questions and finally got Dye on the phone for a discussion, where he learned that when Random House says "own," they mean "not own." In fact, at best, when they say "own" they mean "if you fit into this limited category, you have the right to move your ebooks from one approved platform to another approved platform." This is, contrary to Dye's claim, a license. It is not ownership.

As many surmised, the key phrase in Random House’s communications is “authorized library wholesalers.” In the context of the LJ article, Random House was using a definition of “ownership” that you won’t find in Webster’s dictionary, conveying rights where none exist. In fact, Random will not sell directly to libraries or library consortia, although Mr. Dye reiterated that they continue to evaluate many alternative library business models. RH’s approach in the library market is to vet potential library market distributors for auditing, accounting, security, and other business functions, and then permit libraries to acquire titles from that short list of approved bureaus. In Random’s view, libraries “own” the titles they purchase to the extent that they should be able to migrate their ebook catalogs from one platform, such as Overdrive, to another, such as 3M.

That’s very nice. It’s just not ownership. It’s licensing, with benefits. Library customers of RH titles do not have the ability to transfer their titles to an unapproved platform, such as Califa or Open Library; they cannot resell or donate their ebooks; and there is no mechanism for libraries to receive ebook donations directly from consumers. All that libraries “purchase” from Random House is a verbal commitment to assist libraries in moving their Random House ebooks from one approved commercial platform to another. This is the kind of “perpetual license” that academic libraries have traded for ownership. Academic libraries now employ licensing specialists, and see the world through the lens of contracts. In consequence, faculty have begun to develop open access models that revolutionize scholarly communication from within.

In other words, Random House's claims were a load of bull. That's not surprising, but still disappointing. Brantley goes on to say what kind of ownership should be allowed for libraries when it comes to ebooks:

Public libraries seek a different kind of ownership – the kind that appears in the dictionary. The Internet Archive, Douglas County Libraries, Califa, and a growing number of other library systems are running their own ebook platforms, providing their own auditing, accounting, and security. We want to keep ebooks in our communities, run our own services, safeguard the privacy of our users, and be free from overreaching licensing regimes that threaten our services. And increasingly, we are finding publishers who are willing to sell to us directly, seeing the benefits of handing management of digital titles to libraries. Libraries can market e-books to the people that want them, and gather usage statistics in a privacy-protecting manner to help inform other libraries – as well as publishers – about what titles are popular, and where. These are rights and responsibilities that publicly funded libraries should not hand over to commercial distributors that must navigate between the Scylla of publishers and the Charybdis of Amazon. Readers First is an example of the larger movement articulating libraries’ desire to re-forge a partnership between publishers and libraries.

Of course, I'd also argue that this goes way beyond just libraries. Users want to own their own ebooks as well, just like they own physical books. That means they don't want to worry about having the company they bought their books from suddenly lock them out of their collection for reasons they won't explain. It means they want to be able to move those ebooks from platform to platform without permission. It means they want to be able to lend those ebooks to a friend. Some smaller publishers get this, provide DRM free ebooks, and make it easy for this to happen. Random House, on the other hand, doesn't seem to understand the issue at all.

Recently, Tim noted that, for some strange reason, politicians don't like having the same level of surveillance applied to them as they wish to inflict on the public. Here's a nice case from the state of Uttar Pradesh in northern India, found via Evgeny Morozov, where politicians aren't being given any choice:

UP [Uttar Pradesh] chief minister Akhilesh Yadav has decided to install CCTV cameras at the residences and camp offices of all the ministers in his cabinet.

As usual, the politicians are up in arms:

But a minister told Mail Today on the condition of anonymity that many of his colleagues including Raja Ram Pandey, khadi and village industry minister, Paras Nath Yadav, animal husbandry minister, Awadhesh Prasad, social welfare minister and Durga Prasad Yadav, stamp and registration minister, have opposed it.

The reasoning behind the move is interesting:

'the CM [Chief Minister] found the idea exciting mostly because he could boast of this action during Lok Sabha election campaign to claim that his ministers were observing complete transparency in their activities,' a source close to the CM said.

It's clear from this that there are some ulterior political motives behind this extreme form of governmental transparency. But equally, assuming that Uttar Pradesh's Chief Minister goes through with his plans, there's no denying that it would give politicians a unique insight into the reality of surveillance, and that can't be bad thing.

Indeed, what we need is a general rule that politicians anywhere who are contemplating increased snooping on their citizens must be put under 24x7 surveillance in their homes and offices before they are allowed to enact any such laws. If that were the case, I predict their enthusiasm for spying might well evaporate.

Follow me @glynmoody on Twitter or identi.ca, and on Google+

Calls for evidence-based policy-making in the copyright domain are increasing on both sides of the Atlantic. How do we best regulate the fair remuneration of artists? How do we enforce it? Evidence based on sound methodologies and research is slowly but surely appearing. Now the highly respected Institute for Information Law (IViR) of the University of Amsterdam joins the league of evidence-givers with a new report, Filesharing 2©12, Downloading in The Netherlands, about how blocking websites is not a worthwhile remedy (The report is in Dutch, but the executive summary is translated. It was an initiative by the IViR itself and was partly financed by the Ministry of Culture, Education and Research, some ISP's, Dutch society for professionals in the book industry and done in collaboration with several other institutes. Small disclaimer: I did my masters at this institute).

Economist Joost Poort explains his main findings:

"The most surprising result for me is that the act of downloading from an unauthorized source seems to have peaked and is now declining. This is most likely due to the actions of the music industry, who have been successfully experimenting with new business models."

There you go, music industry! You're doing it right! It seems like everyone is winning in the Netherlands: illegal downloads decreasing, new business models (streaming, advertising) successfully competing with abundant free supply, and paying customers. In fact, the study shows that some two thirds of consumers are willing to pay around the current retail prices for content. I can't back this up, but it seems reasonable to suggest that this is the same amount of people who would have shelled out cash back in the day of optical discs, or CD's as those clumsy shiny discs used to be called. Remember all those thieving fans who proudly owned huge tape collections with self-decorated covers of the copied albums? If you don't, you must be younger than 25. However, not all is good news. Poort continues:

"The downloading of films and series from unauthorized sources is on the rise. This shows that enforcement is not the reason for decreased downloading of music, since successful enforcement would not only have an effect on music, but on all media offered by blocked websites, such as the Pirate Bay"

Note that ThePirateBay.org is blocked in the Netherlands, pending appeals in several cases. Poort elaborates:

"The reactions from consumers show clearly that the blockade of the Pirate Bay website is not relevant to about 75% of them. This is because they never downloaded from illegal sources anyway, or stopped downloading since legal services have taken off or they had enough money to pay for content. The other 25% seems fairly unaffected by the injunction. Only 5% have indicated they actually started downloading less, not a shockingly large effect to put it mildly."."

To say this is an interesting finding is an understatement of epic proportions. This finding gives decisive evidence that blocking access to websites, services or content distribution methods is a bogus way of addressing an innovating customer base. To repeat, but in more joyous terms: Hey Policymakers! Honorable Judges! Here's the evidence you've been waiting for! Ha, turns out that blocking websites is an ineffective way to change consumer behavior after all. Gosh, if only someone had predicted this before... Would've saved us all quite some hassle, right?!

Well then, now this is settled, time to focus all the wasted litigation and lobbying money for restrictive measures on R&D. The legal services are still in their infancy and have so much potential. The better they get; the more people will pull out their wallets to pay for access. You wouldn't want to be the only kid left behind who can't access all the great content made available through 4G network phones and super fast broadband by really easy to use services, would you?

Thinking about it this way, affordable, user-friendly and all encompassing services for music, films and books may even increase the paying user base from 67% to... 100%?! Maybe not quite 100%, but more research is needed for this new and possible social phenomenon of peer-pressure to pay for media and at the same time the reversal of the old media model: huge paying consumer bases for the cultural sector due to abundant and ubiquitous availability, instead of artificial scarcity and thriving "piracy". Record execs must be throwing money at their screens by now, reading about this not-quite-so-hypothetical but under-researched gap in the market.

Of course, we must not forget that part of the financing team is currently in appeal for a court ruling, which forced ISP's to block ThePirateBay.org. This blogger is happy to keep you updated of any developments in the already legendary Dutch copyright fisticuffs of 2012!

Ah, the UN. As highlighted by Declan McCullagh, a new report from the United Nations Counter-Terrorism Implementation Task Force, clocking in at an unwieldy 158 pages (pdf) warns that this old internet of ours is just too damn open, and that means terrorists can use it. Thus, it has to stop the openness. The report really is just about that bad: if terrorists might misuse it, it's bad and must be stopped. The costs of locking up all this openness are brushed aside, if they're even considered at all. Among the problems? How about open WiFi?

ISPs may require users to provide identifying information prior to accessing Internet content and services. The collection and preservation of identifying information associated with Internet data, and the disclosure of such information, subject to the appropriate safeguards, could significantly assist investigative and prosecutorial proceedings. In particular, requiring registration for the use of Wi-Fi networks or cybercafes could provide an important data source for criminal investigations. While some countries, such as Egypt, have implemented legislation requiring ISPs to identify users before allowing them Internet access, similar measures may be undertaken by ISPs on a voluntary basis.

It seems like it should be a general rule that, if you're supporting something that includes better surveillance tools by saying, "Hey, Egypt -- the same country that recently had the people rise up to force out a dictator, who tried to shut down the internet -- does it!" perhaps you don't have a very good argument.

The report is basically one big "OMG! But... but... terrorists! Kill it!" It talks about things like "standardizing" data retention rules for ISPs, while we here in the US don't currently have data retention rules -- nor is everyone in agreement that such things are good. Nevermind all that... terrorists!

The development of a universally agreed regulatory framework imposing consistent obligations on all ISPs regarding the type and duration of customer usage data to be retained would be of considerable benefit to law enforcement and intelligence agencies investigating terrorism cases.

Also... all that social media stuff going on out there? Scary, scary stuff because terrorists might use it as well. They might publish propaganda on it, and we can't have that:

The promotion of extremist rhetoric encouraging violent acts is also a common trend across the growing range of Internet-based platforms that host user-generated content. Content that might formerly have been distributed to a relatively limited audience, in person or via physical media such as compact discs (CDs) and digital video discs (DVDs), has increasingly migrated to the Internet. Such content may be distributed using a broad range of tools, such as dedicated websites, targeted virtual chat rooms and forums, online magazines, social networking platforms such as Twitter and Facebook, and popular video and file-sharing websites, such as YouTube and Rapidshare, respectively. The use of indexing services such as Internet search engines also makes it easier to identify and retrieve terrorism-related content.

You hear that? All those internet companies, enabling terrorists. Oh, and they're not just handy for terrorists to promote their propaganda... but to sneak up on the dumb users who reveal important info for terrorists as well:

Particularly in the age of popular social networking media, such as Facebook, Twitter, YouTube, Flickr and blogging platforms, individuals also publish, voluntarily or inadvertently, an unprecedented amount of sensitive information on the Internet. While the intent of those distributing the information may be to provide news or other updates to their audience for informational or social purposes, some of this information may be misappropriated and used for the benefit of criminal activity.

Loose fingers on Twitter sink ships, as the saying goes.

Now, of course, some of this is just describing what's going on out there for those who haven't realized that any communications technology can be used by both people with good intentions and bad intentions (and no intentions at all). And we shouldn't freak out about that kind of thing. But, the report does also make some "legislative and policy recommendations," where it gets worrisome:

In order to provide effective criminal justice responses to threats presented by terrorists using the Internet, States require clear national policies and legislative frameworks. Broadly speaking, such policies and laws will focus on:

(a) Criminalization of unlawful acts carried out by terrorists over the Internet or related services;
(b) Provision of investigative powers for law enforcement agencies engaged in terrorism-related investigations;
(c) Regulation of Internet-related services (e.g. ISPs) and content control;
(d) Facilitation of international cooperation;
(e) Development of specialized judicial or evidential procedures;
(f) Maintenance of international human rights standards.

Nice of them to throw in that last one about human rights... because all of those other ones are really about ways to chip away (often with a pretty big digital bulldozer) at human rights and civil liberties. In providing examples of countries that have put in place good anti-cyber-terrorism laws... they list a who's who of countries with dubious human rights records, including Saudi Arabia, Pakistan and the United Arab Emirates. Oh, and China:

In the terrorism context, in China there are provisions criminalizing different forms of terrorist activities, including article 120 of the Criminal Law, which criminalizes activities related to organizing, leading and participating in terrorist organizations. This broad criminalization provision covers a wide range of terrorism-related activities, including those carried out over the Internet.

Of course, if you also get to define what counts as "terrorism," I imagine such laws can be quite handy in making opposition parties and activists disappear (or at least get them to shut up).

There's a lot more in the report like that. While the report pays lip service to "human rights" throughout, it really seems to focus on a whole bunch of ways to chip away at those human rights all because terrorists might be out there, using your internets.

We're rapidly approaching the time in which perpetual copyright hits its existing statutory limits -- so I've fully been expecting an increase in arguments for why copyright needs to be extended again. Of course, the actual economic evidence doesn't support this at all. Instead, the evidence suggests there's tremendous value in a broader public domain. So how will maximalists argue for copyright extension? If a recent paper from economist Stan Liebowitz is any indication, it will be through strawmen and the argument that we should ignore the economics. Seriously.

Jerry Brito recently had Liebowitz join him on the Surprisingly Free podcast to discuss a paper Liebowitz wrote entitled Is Efficient Copyright a Reasonable Goal? Liebowitz, if you're unfamiliar with him, has become the MPAA/RIAA's favorite economist over the last few years, and they tend to trot him out in various debates over file sharing. He also seems to have what feels like a personal vendetta against economists Felix Oberholzer-Gee & Koleman Strumpf, who have repeatedly shown how weaker copyrights have benefited society. This particular paper appears to mostly consist of Liebowitz taking some of Oberholzer-Gee and Strumpf's arguments out of context, building a strawman around them... and then arguing that we should ignore economics.

First up, he argues that there are legitimate theorists out there who believe that we shouldn't let artists get too rich -- and that those same people don't seem to mind anyone else getting rich. In the podcast, he focuses on how basketball players get super rich, and we would all think it's crazy to tax most of their income, even if it were shown that they'd continue to play basketball at a lower salary. He notes, correctly, that incomes above the point at which they'd still be playing basketball are a form of economic rents, and the theory is that you can tax such economic rents without doing much harm elsewhere -- and that's a "more efficient" solution. So, strawman number one is this idea that there are people who think that it's okay to stop musicians from getting super wealthy, but who feel otherwise about, say, pro-basketball players.

He argues that there's no way in policy to "tax" basketball players, but because we can control the length of copyright, we can (and do) levy exactly that kind of "tax" on content creators like musicians by taking away their copyright.

We don't actually go after rents in the tax system. No one tries to remove the 90% of rents that people get if they're top basketball players. No one says 'well, gee, Charles Barkley, back in the day, when he was doing it, or Michael Jordan, or anybody now, we should have taken 90% of their income away, because it was all rent.' So the textbook says we should and it would be efficient if we could, if we're just looking to generate revenues for the government. But there's no practical application of actually trying to do any of that.

The one place in policy where we can do something like that, is something like copyright. Because in copyright we can control the length or duration of the copyright, which largely controls the amount of rents that the creative individuals get. Now, it's an average, but that's something that we can do for them, that we can't do for almost any other occupation. And that's sort of the difference here. So we're talking about efficiency in terms of copyright. And what's different here is that the ownership rights that the copyright holder is given is not permanent, it's temporary. And we control how long it lasts by the law. Every other type of ownership, to a large extent, is not considered temporary, and you get to keep ownership as long as it stays in your property or your heirs' property.

Of course, there are all sorts of fallacies in those two simple paragraphs. First, the idea that taking away someone's copyright 70 years after they've died is a form of "taxing their rents" is ludicrous. Nearly every work covered by copyright has exactly zero economic value by that point. The vast, vast, vast majority of such works are not being exploited economically at all. Furthermore, I'm having trouble thinking of economically inclined people who argue that it's reasonable to tax 90% of anyone's income, whether they be basketball players or musicians. While I'm sure there are some people out there who think that megamillionaire musicians earn too much, as far as I can tell, those people feel almost exactly the same about megamillionaire athletes as well. But I haven't seen any serious arguments that we should take away 90% of the revenue those folks make. In part, it's because of the bizarre assumption that Liebowitz slips into his explanation above: that the sole goal is maximizing government revenue. But that's not what anyone is generally trying to do from a policy perspective. If anything, it's quite different. They tend to look at maximizing public benefit.

So, already, we're down a bizarre rabbit hole. How did we get here? Well, not surprisingly, if we look at the actual paper, it seems to come from Liebowitz's obsession with Oberholzer-Gee and Strumpf. He regularly cites that paper as supporting this strawman. Yet if you read their actual paper, it does not (as Liebowitz claims) "imply" that it's socially beneficial to pay creators less. Instead, all that Oberholzer-Gee and Strumpf point out is that there are many reasons why people create outside of copyright, and that a weakening of copyright does not, necessarily, harm the incentives to create, but might provide significant additional benefits to the public.

In the podcast, Jerry Brito valiantly pushes back on some of the more extreme arguments that Liebowitz makes, pointing out that claiming that putting works into the public domain is the equivalent of taking away someone's high salary is making the classic mistake of arguing that copyright is no different than scarce property. It descends into the same old argument, where Liebowitz claims that Brito is arguing in favor of "stealing" and pulls out the silly argument that his position would apply equally to "stealing" a car. Brito points out that this is untrue, and his concern is that copyright, unlike traditional property rights, is actually a restriction on what others can do with their own private property (i.e., you can't make a copy of a work whose copy you already hold). Liebowitz refuses to concede the point, other than the minor concession that non-rivalrous goods are somewhat different than rivalrous ones. But, of course, that difference is everything... and Liebowitz pretends it's insignificant.

Jerry Brito (JB): Let's say you wrote a book, and I want to use my printing press to make a copy of it. I can't do that.

Stan Liebowitz (SL): Right. Not without getting permission.

JB: Precisely.

SL: Absolutely. That's what copyright does. But that's essentially a property right. You can't come and use my car without my permission. That's what property rights do.

JB: Sure, but the reason we might want to cut off the rent in the copyright space, where we just want to cut it off where they've earned enough to induce them, whereas we wouldn't want to do that for basketball players, is simply that, in the copyright space, you're limiting what I can do with my property. My physical property. My printing press. I can't use my printing press to make a copy of your book. It's affecting me. Whereas Michael Jordan performing well, doesn't really affect me one way or the other.

SL: Well, you can print anything you get the rights to print. I don't think it affects your freedom any more than it restricts your freedom to drive my car whenever you want to.

JB: No, but when I drive your car, then you can't drive it. But if I use my printing press to make copies of your book, you still have your book.

SL: I no longer have the earning capacity that I have from writing the story.

JB: Which is why we have copyright. And the question is at what point should it end?

SL: That's right. And you're talking about the non-rivalrous consumption aspect of intellectual property.

JB: Exactly. At what point should the limit on my freedom be lifted?

SL: Because it's an issue where you want to have the freedom to generate revenues from an item that I created. That's one aspect of it. I'm not sure we should give you that freedom, just as we shouldn't give you the freedom to sell my car. I mean yes it's true that one's rivalrous and one's not rivalrous. But I create the story. Why should you be allowed to generate revenues on a story that I created?

JB: I guess I'm...

SL: What does that have to do with your freedom? We do view the property right differently on copyright than we do on other items, but I'm not sure we should. You claim there's a difference in freedom and I'm saying, 'no, just as you can't take my car without my permission, I don't see why you should be able to take my story without my permission.'

JB: Let's not get too deep into this, but I'd simply say...

SL: I mean you want the freedom to steal, is what you're saying.

JB: Well, no...

SL: It infringes on your freedom.

JB: No, no, no...

SL: If you consider an item, and we'll say let's treat it like other property, then yes, my story is my story. And you want to steal it.

It continues in that vein. Perhaps for way way too long. It's not hard to pick out the series of logical fallacies that Liebowitz relies on there. He doesn't differentiate between actually taking something (such that the original person no longer has it) and making a copy of the work, such that there are now more copies in the world. Those two situations are extraordinarily different both conceptually and economically. He refuses to recognize the simple fact that the very basis of copyright is restricting the freedom of others (perhaps for legitimate reasons -- but restricting them nonetheless). He doesn't seem to recognize that the ability to benefit (profitably or not) off the works of others is not a horrible thing -- and, in fact, it's a key piece of how economic growth works. That is, Liebowitz appears to be arguing against economic growth, by suggesting that all externalities are inherently "bad," contrary to nearly all economic literature on the subject.

He also tosses out the ridiculous paired fallacies that (1) ignoring the copyright takes away all "earning capacity" from a work and (2) taking away earning capacity represents a real harm or loss. Neither is true. There are still plenty of ways to benefit without copyright, and uncaptured earning capacity is a marketing problem. It means you failed to convince someone to pay. That's it. If Liebowitz's argument that harming someone's "earning capacity" is somehow an unfair removal of their property rights -- which is exactly what he claims above -- is accurate, then he would appear to be against all competition. That is the very nature of competition, of course: to try to take away a competitor's earning capacity, such that the revenue that might have gone to a competitor goes to you instead.

In the paper itself, thankfully, Liebowitz's argument is somewhat more thought out and nuanced, though still not at all convincing. He shows the traditional curves of demand vs. marginal cost and marginal revenue -- and the price and quantity produced under the different circumstances, arguing that:

Under a regime of competitive production of reproductions (no copyright), the equilibrium price of a reproduction is Pc (equal to marginal cost, MC) and the quantity of reproductions purchased is Qc. There is no profit earned in the market for reproductions since the market is assumed to be competitive.

However, as we've discussed for years, this simplistic analysis assumes a static market, and one where the producer does not innovate or provide differentiated, scarce value above the product itself. That is, musicians can (and do!) create additional value that they get people to pay for that can't be copied, such as the personal connection between the artist and the fan, a live show, physical merchandise, the beneficial feeling of supporting an artist, etc. In other words, the classical analysis is way too simplistic, and ignores how everyone, whether dealing with rivalrous goods or non-rivalrous goods, prevents price from being pushed to marginal cost: by adding additional scarce value. The problem is when you focus solely on the non-rivalrous good, and assume that there are no additional scarcities (tangible or intangible) that can be created around it. While Liebowitz does make a nod to this fact later in his paper, he doesn't spend much time on it.

The end result of all of this? Liebowitz more or less admits that his argument does not make economic sense, and then still seeks to justify perpetual copyrights.

Let's remember that the key point I'm trying to make is one that is not particularly amenable to economic analysis. I don't deny that it's efficient to sort of weaken copyright to a certain balancing point. What I'm saying is that we can do that, we have the tools to do that, and no one finds it particularly morally objectionable to weaken copyright to get to what is the proper balance. The point of my paper is that if we were doing it the way we're saying we would like to, we would be removing rents from the creators of copyrighted works, because that's how you get the balance that's most efficient. But we don't go reducing rents elsewhere in the economy, say from basketball players.

In other words, the argument here is: because basketball players are rich, we shouldn't reduce copyrights, because then musicians might not become quite as rich.

Of course, those arguing for increasing taxes on the rich might disagree about the claim that no one wants to "tax" the super rich in other fields. But the larger point here remains... well... confounding. Liebowitz makes a number of assumptions that simply don't make economic sense -- and even then admits that his conclusion doesn't make economic sense -- but still suggests that perhaps copyright should be perpetual... because, in his view, we don't tax other professions.

There are still a few more years before the real battle over copyright extension heats up, and I fully expect that we'll be seeing some crazy arguments between now and then for why copyright should be extended, but if this is the best that the MPAA and RIAA have got right now, they may want to seek some additional help.

Back in August, Mike wrote about some questionable sharing of license plate information between the US Border Patrol and various insurance companies. While the stated aim of tracking stolen vehicles might seem to make this sharing justified, the fact that this is going on with no oversight or accountability is cause for alarm.

Of course, law enforcement has long had big plans for license plate readers (LPRs). The ACLU has come across a recording of a 2010 National Institute of Justice conference in which one of the speakers expresses an interest in building LPR data into a "Google" of license plate locations.

Dale Stockton, Program Manager of the “Road Runner” project at the Automated Regional Justice Information System in San Diego spoke on a panel on license readers at the 2010 conference and explained to police and prosecutors in attendance how best to share license plate data. Mind you, he was talking about the location information of people never accused of any crime.

Stockton knows this would never fly if attempted directly. He states as much in the transcript:

We're probably not going to have any centralized national giant bucket of license plate reader data. It probably wouldn't stand the court of public opinion, and it's probably something that, given where we are in the rollout cycle, wouldn't easily be done, but we can develop regional sharing capability...

But the "court of public opinion" can be routed around, according to Stockton. Despite frankly stating that the public would find a "Google of license plates" odious, he intends to do just that, through a series of back doors.

And so doing, you get those set up and then begin to share between those regions, and as you begin to look beyond your region, utilize a trusted broker like Nlets...

Every law enforcement agency has a connection to Nlets. Nlets would serve not as a storage unit but as a pointer system, something akin to a Google, so that when you check a plate, Nlets would point you in the direction of where that plate can be found, and the result of that would be a query in one state by an investigator could give an indication of plates of interest in other states, and then that information can be pulled back.

If Stockton has his way, a decentralized search system for license plates, routed through a third-party's software, will perform exactly the way he wants it to. Somehow he feels that a distributed system is OK while a centralized system isn't. Or rather, he feels that both systems are OK, but the public will only put up with the illusion that law enforcement isn't running a Google-esque system of harvested plate data.

Is the fear of a system like this overblown? Every law enforcement official at the conference runs down several anecdotes about how the plate reading system has aided them in investigating various crimes. But should law enforcement have access to nationwide plate data, much of which pertains to citizens who have never been accused of a crime, much less committed one? Stockton tries to justify the LPR system by comparing it to officers running plates in person.

One of the questions about license plate readers is this kind of hocus pocus, "you are invading my privacy; it's super intrusive." And I come from the opposite end of the spectrum. I truly believe that this technology is only doing what license plate readers have already — I'm sorry — what officers have already been doing in the field for many, many years.

The courts have indicated to us that officers can look at a plate and run it. So let's just go down quickly here through the two sides, an officer doing it and the license plate reader doing it. Officers can run plates anytime they want. The courts have held that's why we put the state plate on there, so an officer can check and make sure it's current and it's not wanted. Well, the license plate reader is doing the exact same thing. It's looking at it; it's checking that plate to see if it's wanted. The officer can pick and choose among the vehicles that he or she looks at.

The biggest difference here is the "always on" aspect and the fact that everyone is tracked. "Reasonable suspicion" and the like are taken completely off the table and replaced with an indiscriminate system that harvests data. It's a handy way to peel back another layer of privacy, as Kade Crockford of the ACLU states:

We've been making a lot of noise about location tracking of late. License plate readers rank high among the technologies that are threatening our privacy with respect to our travel patterns. Where we go says a lot about who we are, and law enforcement agencies nationwide are increasingly obtaining detailed information about where we go without any judicial oversight or reason to believe we are up to no good. Stockton says we have nothing to worry about with respect to license plate reader data and privacy, that that's all "hocus pocus." But he's wrong.

We must ensure license plate readers do not become license plate trackers.

As Crockford points out, the Department of Justice has already stated that Americans have "no privacy interest" when it comes to cellphone location information. The DOJ's rationale could easily stretch to fit this scenario. But what's "good" for the American public is rarely viewed as acceptable for those writing the guidelines or using the technology. For all the talk about "nothing to hide," officials are very touchy when it comes to making their information accessible.

Data from license plate readers in Minnesota was obtained by a St. Paul car dealer using open-records laws, and used to repossess at least one car, according to a recent article in the Minneapolis Star Tribune. The article included this amusing tidbit:

"When the Star Tribune published data tracking Mayor R.T. Rybak's city-owned car over the past year, the mayor asked police Chief Tim Dolan to make a recommendation for a new policy about data retention."

So, the question that needs to be asked of every politician and law enforcement member who feels this system will only be used for catching "bad guys" is whether or not they'd mind having their location tracked via license plate readers. Stockton's pushing for something he knows the public won't stand for and is using successful investigations as the ends to justify the privacy-violating means.

Ah, stupid copyright licensing rules block perfectly normal activities yet again. This past weekend, Saturday Night Live ran a mildly amusing skit involving a power outage at internet streaming radio company, Pandora, in which an intern -- played by Bruno Mars -- has to step in and sing a variety of songs to keep the streams running. It's a slightly hacky trick to show off Mars' singing mimicry, but done pretty well. While NBC has had a somewhat ridiculous love/hate affair with putting SNL clips online. Over the past few years, it's finally realized that viral clips are an important promotional vehicle for the show. Yet... this clip is not online on NBC.com or Hulu, where SNL normally puts its clips... because (of course) music licensing online makes it an impossibility. The TV shows have licenses for TV broadcast, but they don't apply to internet streams (which is why you see some shows change out their music on Hulu). Yet, here, the clip doesn't work at all without the actual music.

Of course, this is the internet, so the clip was quickly uploaded all over the place, and while some of those sources have already seen it pulled down, others still seem to have it up. At the time of this posting, eBaum's World appears to have a working copy. Of course, having now seen it, it would seem like there's a pretty strong parody defense claim if anyone argued they were infringing. Mars parodies many of the songs, changing or garbling the lyrics, which is a big part of the joke. But, of course, that would require NBC Universal to actually have the guts to fight in court for fair use -- and even just thinking that I think I heard some laughter coming from Rick Cotton's offices.

The end result, though, is nothing but stupidity. NBC doesn't get to show the clip more widely and get the promotional benefits. It also doesn't get the ad revenue that would have gone alongside its own hosted clips. Instead, other sites get the attention and the traffic.

Following the recent pro-fair use rulings concerning Georgia State's e-reserves and book scanning efforts by a bunch of colleges and universities in the HathiTrust case, Siva Vaidhyanathan has a good article pointing out that these two rulings recognize that universities are "vast copy machines" and that's a very good thing.

...these cases strengthen the claim that universities and their libraries have a special place in copyright law because they have a special place in society. Courts and even Congress have long acknowledged the essential role of copying in the educational process. That’s why the preamble to the section of the 1976 U.S. Copyright Act that outlines “fair use” specifies “teaching (including multiple copies for classroom use), scholarship, or research” as examples of “fair uses”—uses that, although they involve the copying of protected material, are considered noninfringing because they enable essential public goods.

Universities are not copyright-free zones—far from it. But they do perform special services that often demand flexibility and liberties that enable them to “promote the progress of science and useful arts,” the core mission of copyright as declared by the U.S. Constitution.

I actually think that Siva underplays the importance of learning in copyright -- as that was its original intent. The Constitutional clause concerning copyright talks about "promoting the progress of science" -- which at the time it was written meant "learning," and the very first Copyright Act in the US was actually entitled "An Act for the Encouragement of Learning."

That said, I'm a bit disappointed that Siva then takes that storyline even further to suggest that universities deserve special protections against copyright enforcement that should not apply elsewhere. For example, he insists that while the HathiTrust ruling finds fair use for universities in scanning books (even if the scanning is done by Google) it should not apply to Google scanning books itself:

In any event, we are never going to see that operatic courtroom showdown between Google and publishing industry over whether the sweeping scanning of millions of books for explicitly commercial purposes constitutes a fair use by Google. I have written many times over the years that I am dubious of the strength of Google’s argument, and nothing in either the settlement news or the HathiTrust case has undermined my conclusion.

... Google is not a library. It is not a university. It is not a public service. It is a business. Too often we forget those distinctions. The project of creating, maintaining, and offering vast collections of digital material should be something that universities and libraries control, not something we depend on one company to handle.

I have long argued exactly the opposite, that Google's book scanning project should absolutely be seen as fair use as well. The use is clearly transformative and most of the evidence suggests it increases, rather than harms, sales. But, more importantly, it really is creating a tremendous library index that will help people find information and learn. I think Siva is going way too far in suggesting that universities should get special treatment. People learn from all sorts of places, not just universities. And we shouldn't carve out special rules for universities that, by default, harm those who do not attend or cannot afford to attend those universities. Part of the wonder of something like Google's book scanning project (and other similar projects, such as the one by the Internet Archive) is that it helps make these works accessible to all -- and we should be encouraging that for the sake of learning across the board, rather than just for a few institutions.