O'Reilly Digital Media

NB: This is a complete rewrite of this blog entry for the sake of readability.

The lurking question for anybody who opposes compulsory licensing is what alternative they have to offer. I vacillate about whether I’m for or against compulsories, so I’m going to put my alternative to the test by articulating it here.

The most common form of the idea of compulsory licensing is that rights holders are compelled to make their content available, listeners are taxed, and the revenues are distributed according to the results of some kind of census. I advocate a different kind of compulsory: rights holders should
be forced to get off the fence — they must sell their music online –
but how they do it is up to them. There would be no global pool of
revenues. Rights holders would choose their way of selling and
listeners would choose their way of buying.

This can be done by eliminating statutory damages for file sharing.
What are statutory damages? From href="http://www.gigalaw.com/articles/2000-all/landau-2000-10-all.html">GigaLaw:

They may be awarded to the plaintiff even though the
actual market harm may be quite small, or difficult to calculate with
certainty. Although technically not “punitive damages,” in the
traditional tort sense in which damages can be designed to punish
someone, they are meant to send a strong message, “DO NOT
INFRINGE!”

If you get hauled in to court by the RIAA, they’re not going to sue you for money they actually lost (monetary damages). They’re going to sue you for $150,000 per song, regardless of how much they made or lost (statutory damages).

The deadlock

Online music sales haven’t taken off because downloaders can’t buy major hit songs at any price. There are other reasons, like overcharging and underdelivering, but those can work themselves out; if DRM makes a
download unusable people can choose not to buy it, for example. This problem isn’t working itself out because of a deadlock where
rights holders are waiting to see what happens, while buyers are
waiting for hits to become available.

Eliminating statutory damages breaks the deadlock by forcing rights holders to put their songs up for sale. That would create a virtuous circle where buyers can find what they want, which brings them into for-pay systems, which generates revenues for sellers, which gives sellers a revenue base on which to sue for monetary damages.

A practical design

This solution is minimally burdensome to rights holders because it
relies on the free market. Sellers can opt for DRM or not. They can
sign with streaming services. They can sell via their own web site.
They can do whatever they want to as long as they do something.

Unlike designs where there is a pool of revenues which are paid out according to a census, this system would have very little administration. There would be no census bureau and no rights collectives. There would also be less corruption, since there would be no monitoring system to game.

Unlike designs based on tipping, this system doesn’t require
revolutionary change. Rights holders would do more of what they’ve always done — finding and negotiating with distribution and marketing
partners.

Lastly, eliminating statutory damages is politically feasible, since
it can be enacted in a pinpoint bit of legislation. One line in a
rider and the thing is done.

Fairness and freedom

Sticking strictly to monetary damages helps to balance the quid pro quo of copyright. Exclusive rights are a bargain between the public and the rights holders. We gave them the right to sue 12 year old girls into oblivion in exchange for making their music available, they have to live up to the bargain.

Eliminating statutory damages would also fix the problem of gross
unfairness towards people being sued. It makes no sense for Brianna
LaHara’s mom to be on the hook for $150,000 per song times 1,000 songs
– $150,000,000! Exclusive rights are intended to be a stick, not
a murder weapon.

In terms of freedom, eliminating statutory damages is the least
restrictive solution on the table. The only behavioral restriction is
that musicians who want to sue have to put their music up for sale,
which is a reasonable requirement regardless. There is no new
regulatory body. It’s fully decentralized, except that courts will be
involved in deciding monetary damages.

Advocates of freedom will take
a harder line, saying that intellectual property is evil. Fair
enough, except that freedom is comprised of our actual state, not just
the one we see through the filter of the law. Eliminating statutory
damages leads to an actual state that’s more free because it’s more sensible. Somebody getting sued by the RIAA won’t walk into court with 99% of the outcome already decided.

Wrapping up

Getting rid of statutory damages doesn’t mean getting rid of monetary
damages. It means saying that somebody can’t sue for money they
haven’t lost. If they can’t sue for money they haven’t lost, they
have to be in business, and if they’re in business — if they are
really making their music available to the public over the internet
for a plausible price — then the market will be able to work its
blundering magic.

Don’t get me wrong, as a ubiquitous client, the
web browser is hard to argue against.  Despite many pitfalls, it’s got
tremendous momentum.  At the same time, I have been thinking.  When
do client applications matter?  Simply put, when does a rich
client have an "unfair advantage" over a web browser in
terms of bringing real value-add to the application composite?  

In thinking this one through, seven application patterns
stand out (I am sure there’s plenty more), and it seems clear that there is
lots of potential synergy between these attributes. style='mso-spacerun:yes'>  Needless to say, this is something to think
about as you build your next generation Internet applications: 

1. The application leverages a
   local data store.
2. There are lots of
   associations between items in the store.
3. The application relies
   on near-real time filtering to provide customizable user views.
4. Maintenance of
   a transparent information flow between a user’s other
   client applications is integral.  
5. Reliability of data access is
   mission critical to the application. 
6. Multimedia content
   is regularly presented by the application. 
7. Peer-to-peer communications
   or shared work spaces are a fundamental part of the application.

Do you buy my logic? Do other scenarios stand out for you?

The business of music is making hits, which generate exponential returns on investment. The business of gaming a compulsory licensing scheme requires rigging one vote at a time, which generates linear returns. Because hit making is a better business than vote rigging, a compulsory licensing scheme is more likely to be oriented towards music than corruption.

………………………………….

This is a proof that gaming of compulsory licensing schemes is a non-fatal problem as long
as output dollars to rights holders are no greater than input dollars from
fees or taxes.

Allow that there are open auctions for those votes available to purchase.
Then the output value of a vote minus the input cost of a vote will tend
towards zero and profits will rise linearly but slowly as a multiple of
that ever smaller amount.

But popularity follows a power law curve, while the upfront costs of
attaining it are only linear, so profits from popularity would grow
exponentially rather than linearly.

In other words, it is more profitable to pursue extreme popularity than to
buy votes. It is not an accident that the CD industry lives on hits, and
it is not likely that a music industry funded by compulsories would change
from a revenue model primarily based on hits to a revenue model primarily
based on vote buying.

………………………………….

Attacks on the monitoring side of a compulsory licensing scheme will always be a worse business model than simply producing hit records. The music business exists as it is for a reason.

Comments

Neil Netanel responded, in a post to the Pho list:

But, and here I will borrow from the Cyberprof discussion, the gaming
problem is not limited to the music industry. What if a political
activist group, let’s say the NRA, decides to game the system as a
means to raise money. It produces 200 low-quality, cheap pro-gun
“songs” and asks its tens of thousands of members to download and play
each as a “donation” to the cause. If these are real songs, I say –
great. Let’s have more expressive diversity. But what if each “song”
file is really just a hastily created, low level employee humming I
like guns into a tape recorder — that even NRA folks wouldn’t really
listen to. Would it be worth it for the NRA to do that? Or would it
make more sense for it to invest in catch songs that might galvanize
supporters?

The NRA would make more money by generating popular songs.

Its members who offer to help by downloading the songs would be foregoing the chance to sell their votes elsewhere at full price, and full price would approach the exact value that the NRA would stand to gain, so the game would earn no more than if the NRA solicited their contributions as cash.

The NRA would only make a profit by generating songs whose popularity goes beyond direct supporters, which means that the integrity of the compulsory system would not have been compromised.

Related reading

My thinking here was inspired by Andrew Clausen’s paper How much does it cost to buy a good Google PageRank?. Clausen’s proofs are with regard to directed graphs of the kind that PageRank measures, which a monitoring scheme for compulsories wouldn’t likely be. However his
insight that the PageRank algorithm pegs outputs (pagerank scores) to real input dollars (the cost of a domain name) is very relevant. (I didn’t understand the proofs in this paper, by the way.)

Netanel pointed to Peter Eckersley’s proposal on how to protect against ballot stuffing, Virtual Markets for Virtual Goods.

Chris Grigg points out that Aaron Swartz’ Fixing Compulsory Licensing is a method for implementing a vote mechanism. He’s right. Aaron says: when you pay the tax you get a vote. That’s the same as here. Where this is different is in this bit from Aaron’s essay: There is a chance that everyone will give all their money to themselves, but this can be prevented by only paying out to accounts that meet some higher threshold of cash. The mechanism I propose here shortcircuits that kind of vote rigging by allowing users to sell their votes. If they can sell their votes, giving them away — even to themselves — means losing money.

Okay, I admit it. I am really passionate about the consumer space. It’s interesting to me to imagine how much gravitational pull AOL once had, and what they have now. To ponder how Yahoo becomes the next AOL. Of course, I am always asking myself where Microsoft settles in online. On different days, I can argue that the browser will fade away or that it will become more ubiquitous. Is Google unstoppable?

In contemplating the above, and particularly how the “old” web (of web pages and web sites) morphs into a new and improved Internet, there are four general trends I plan on writing about in this blog. I definitely don’t come from the perspectives I will put forth from on high, so challenge my assumptions and pick them apart, for this space is really just a straw man for further discussion.

“Grope. Ship the idea. Fix. Iterate.” That’s my motto, and that’s what I’ll attempt to do in subsequent blog posts relative to the following:

1. More intelligent search: Once upon a time, search was undifferentiated, and finding the information you were looking for was a function of plugging in the right search terms. But search is evolving, and becoming more intelligent. In the process, both the types of information distinctly indexed and supported (think: web sites and web content, local business listings, news stories, products and comparison data) and the applications of what you can actually “do” with the information returned will evolve meaningfully.

2. Software as service, and service as software: Once, there were standalone applications. Then there was client-server. Then the web browser became the ubiquitous client, and everything else seemed so 1993. But, as concepts like integration, persistency, reliability, context, community and mobility become of greater importance, expect to see Internet clients evolve, muddying the question of when software is merely a service and when a service is integrally bound to software. I’ll attempt to storyboard some different application scenarios, and the role of the different frameworks and standards in facilitating them, and play the devil’s advocate in terms of when web services make sense for the provider and when they don’t.

3. Channel me, or my personal online truth: A recent study by MasterCard International asserts that nearly one-quarter of the online population is “Confident Core Users.” This group has been online a number of years, is comfortable buying products online, has broadband connectivity and is utility seeking in their utilization of the Internet. As the emphasis of consumers shifts away from technology for its own sake, and towards online productivity, new models and new applications will need to emerge that enable users to better manage their recurring digital lifestyle activities, topics of interest, and the people, products and businesses that they encounter along the way. Among the open questions on this one is where does all of the information accumulated actually “live” (think: IM sessions, email, IM, RSS feeds, blogs, web content, listings) and what is the front end or front ends for tying it all together? And finally, how does wireless and devices in the smart, connected living room factor into the equation?

4. Online spaces: The concept of the Internet as an online “space” is nothing new, touching both specific types of applications (think: blogging, chat, email) and specific types of activities (think: dating, marketplaces, education). What’s changing is the structure, boundaries, reach and transparency of such spaces — from clear purpose and clear boundaries (think: eBay, Amazon, Craigslist) to virtual boundaries that are defined in real time and provide transparent linkages to six degrees of separation (think: Friendster, Tribe.net). Is this trend the next ICQ, is it another Hotmail, or is it destined to be the next PointCast?

A final note aside. In writing about the above trends, where practical, I will speak not only to the “how” and “why” sides of the equation (i.e., how such and such is built and why a given approach is compelling), but what’s the business model to support it and the various and sundry human factors that stand in the way of the scenario playing out. Should you be interested in such things, I have a set of constructs that I follow in sniffing all of this out called “Core Concepts Learned in The Internet Economy,” which is basically a bunch of analogs borne of firsthand experience in trying to reconcile both the strategy and tactics of building compelling products and a vibrant company. As a participant in this blog, you are welcome to download it.

What’s your take on where the consumer Internet is headed?

Related link: http://www.skype.com/

P2P telephony from the kazaa guys.
I hear that they proxy the phonecalls for the double nat problem. Interesting.

…

Related link: http://www.geourl.com/

GeoURL is a location-to-URL reverse directory. This will allow you to find URLs by their proximity to a given location. Find your neighbor’s blog, perhaps, or the web page of the restaurants near you.

Related link: http://www.mapbureau.com/rdfmapper/

RDFMapper is a web service that searches an RDF file for resources with geographic locations, and returns a map overlayed with dots representing located resources. Clicking on a dot displays a web page representing the clicked resource. Arbitrary images can be treated as maps, so the service can be used for any kind of image annotation.

Related link: http://www.mcs.vuw.ac.nz/~chikken/research/writing/chinz_paper/

A wordy but good academic/intellectual look at how metaphors are used in user interface.

For a while now, I’ve been watching from my editor’s desk about the
convergence between film and digital moviemaking. And if you’re like
me, you probably didn’t see many movies in the theatres this summer
that made you stand up and say, “Wow! That was a solid movie!”

Well, neither did Hollywood–except they’re referring to the
bottom-lines at the box office. So here’s an innovative way for
Hollywood to cut costs in making theatrical releases: give college
students a million dollars, have them develop a script (maybe), edit it
together, put in some FX, and stick it in a theatre near you. See this
article in the Austin American-Statesman newspaper for more details.

Curtain Rises on UT Film Institute

This is interesting for at least three reasons:

1. It will be entirely digital productions, hence an increase in the
   amount of people learning affordable PC and Mac production tools such
   as Final Cut Pro, After Effects, Premiere, Audition, etc.
2. It’s blessed and funded by Holywood studios with plenty of big
   wigs to ensure that distribution actually happens into major theatres
   (UA, Cinemark, General Cinema, etc.)
3. The test bed is happening right here in Austin– my own back yard!

As I mentioned, this isn’t an entirely unexpected move, given how many
underperforming big-budget movies there were this summer. Many of them
were good, but their budgets were too high, and only some squeaked out
with a small profit domestically (numbers courtesy of imdb.com).

For example:

Terminator 3: Rise of the Machines
Budget: 170 million   Gross: 148.4 million (8/24)

Lara Croft: Tomb Raider: The Cradle of Life
Budget: 90 million   Gross: 62 million (8/24)

The League of Extraordinary Gentlemen
Budget: 78 million   Gross: 64 million (8/24)

Hulk
Budget: 120 million  Gross: 131 million (8/24)

Charlie’s Angels: Full Throttle
Budget: 120 million  Gross: 99 million (8/24)

Seabiscuit
Budget: 86 million   Gross: 92 million (8/24)

S.W.A.T
Budget: 80 million   Gross: 82.8 million (8/24)

Bad Boys II
Budget: 130 million  Gross: 132 million (8/24)

And this summer’s favorite whipping boy…

Gigli
Budget: 54 million    Gross: 5 million (8/10; removed
from theatres)

Only a few clear big-budget winners this summer:

X2: X-Men United
Budget: 110 million  Gross: 214 million (8/24)

Finding Nemo
Budget: 94 million   Gross: 329 million (8/24) (Highest
summer movie)

Pirates of the Carribean: Curse of the Black Pearl
Budget: 125 million  Gross: 281.3 million (9/7)

And then there’s this movie, shot with the same prosumer camcorder that
I have, the Canon XL1s, and blown up to 35mm film from miniDV (720 x
480):

28 Days Later…
Budget: 9 million*    Gross (USA): 43.5 million (8/24)
   * Distributed in the US under “Fox Searchlight”,
     which helped finance the film.

So, when you look at the fact that this amateur movie grossed 43.5
million dollars for Fox, the ideas in the article above makes more
sense. Or cents. Take your pick.

Don’t shelve that prosumer camcorder just yet….

Related link: http://shirky.com/writings/fame_vs_fortune.html

Clay Shirky lambasts BitPass, a new micropayment system designed to let content creators set a price of 25 cents or less for a piece of content on the web. Shirky argues that people don’t want to think about how much a piece of information is worth:

First off, you can only truly judge the worth of a piece of content after you’ve ingested said content — not beforehand. Second, the general populous has enough problems with fractions as it is, so how can one expect people to judge miniscule values everytime they want to read some piece of content on the net?

I agree with Shirky that micropayments for content online are doomed to fail. However, let’s turn the tables and instead of an artificial scarcity based model, consider a gift economy model. Micropayments in the gift economy model do make sense — after you’ve read a piece of content you have a feel for the value of the content. This makes it much easier to determine if a piece of content met your expectations and whether or not you should donate the suggested donation for the content.

Furthermore, in a gift economy every participant is equal. Every person can be a giver or a receiver — BitPass makes a distinction between the two and has spenders and earners. While the website is in Beta it looks to be easier to become a spender than an earner, and therefore the two participants are not equal. I think the only difference between an earner and a spender should be reflected in their actions.

Finally, Shirky argues that micropayents present unreasonable mental transaction costs — if you’re constantly thinking about the worth of a piece of content in terms small fractions of a dollar, it does get pretty tedious. In a gift economy the content consumer can decide how much to donate to the content producer, since the consumer is not held to some arbitrary price. I can see creating systems where the mental transaction costs can be shrunk to a small one time mental transaction cost.

For instance, when the user sets up their gift economy based micropayment account the user should be able to set a standard donation price for various pieces of content. A weblog entry could be 1/100th of cent. An image 1/10th of cent. An article a nickel. A song $.25. A movie could be priced at at $1. Whatever the user decides.

After the user ingested piece of content, the software/website/whatever could ask the user: “Was this article what you expected? (Totally, mostly, marginally, not at all)” If the user responds with totally the entire predetermined amount gets donated to the content creator. 66% for mostly, 33% for marginally and nothing for not at all.

The mental transaction cost of this is minimal. The consumer will instinctively have a reaction to a piece of content and answering one question is drop dead simple.

So, I disagree that micropayment systems across the board will fail — there are useful models. However, paying for content before you’ve ingested said content, is not one of those models.

What other uses for micropayments do you see?