Note: Post updated 20/July/2015 to correct factual errors about Ascribe. For more information please see this.
Twenty-five years on Internet is still struggling with the issue of rights management. Existing Digital Rights Management (DRM) systems are unfriendly and technically wobbly.
One approach that is almost comical in its simplicity is something like “Frame” by Depict. Seeking to be the “iTunes of Art”—which we don’t actually want, thank you very much—they envision a large collection of digital objects (images, possibly other things), and a captive platform to deliver it.
There is a certain logic to this. To enjoy visual art digitally, you don’t want a small screen, you want a BIG screen. And since it’s “art”, it needs to have a decorative frame!
I give you “Frame”.
Sharaf, X-Ray, 2014
My own view is that this product is doomed. Too expensive, too difficult to monopolize the content. And probably eminently hackable, if anyone cares enough.
They should sell (a) software that works with any large screen and (b) “frames” to decorate any screen. The software concentrates on tracking the content, and maybe the people who own it. (Require biometric authentication to download content?)
Put those underemployed picture framers to work, installing custom frames on all those ugly HD TVs people are buying. Now that might be a viable business.
Ascribe “Ownership Layer”
“It really is about ownership and understanding what is going on with work that’s attributed to you. That’s kind of where we are,”
Their white paper gives their analysis of the problem and their solution.
They tackle the problem of “how to collect digital art”. How do you sell unique or limited editions of digital objects, and how does one possess such an artifact? They note the widely used workaround, “to try to induce physicality in some way”, e.g., via a signed certificate or other non-digital token.
While I have no interest in collecting digital art, the same ideas apply to many other important domains. In this new age of making, personal digital fabrication (such a 3D printing) allows the easy sharing of executable designs —which is sure to wreck havoc on conventional IP systems, and also thrusts each of us into the messy world of “ownership” of digital artifacts.
Another issue is the question “where is my stuff?” How can I track usage of digital stuff, in order to impose a variety of rights. Currently, the World Wide Web has two settings, “not shared”, and “open to anyone”.
They note that this all came about because of a design decision in the early days of the WWW, the choice to use unit-directional links. They are totally correct in this assessment.
Historical note: I was there. We were aware that there are quite a few variations on what kind of “link” relationships are useful, and the WWW unidirectional link is the simplest of them all, especially since they are also unreliable, “I think what you want is there”.
Bidirectional and even more complex relations are not only important for “ownership”, they are also really useful for indexing and searching. Google would not exist if we had two directional links, because the “page rank” algorithm is nothing less than a way to simulate bi-directional links.
There were basically two reasons why uni-directional links were selected. First, bi-directional links generally require some kind of registry, and the WWW was intended to be decentralized. Second, implementing uni-directional links is easy and cheap, bi-directional are much more difficult. Basically, the decision was to implement something easy that would scale up very rapidly, even if it was no where near ideal, theoretically. The rest is history.
Later, the Semantic Web was developed to implement additional linking schemes. Indeed, the SW can support arbitrary linkages (not just “owned by”, but also “ownership certified by X”, and “X is trusted by Y is trusted by Z”, and I trust Z). Note, though, that the SW has been really HARD to implement.
Ascribe’s (sociotechnical) solution has several parts that coexist with regular web content (i.e., do not require everyone to adopt new software). “The approach has two pieces: a registry with easy, secure legals; and visibility into usage / provenance of the content.” They use the Bitcoin blockchain as a key part of their solution. A second component will be large scale Internet search that discovers similar images.
Unlike many blockchain applications, Ascribe is not trying to create a “trustless” system, nor one that is outside the legal system. In fact, they have done careful work to make it easy to make legally sound claims, which are recorded in a (conceptually) central registry. Done right, this is the foundation of identifying unique patterns of bits, i.e., digital objects.
(One can wonder if the “legals” from Germany are recognized, honored, and enforced everywhere. At a guess, I’d say “no”.)
The Bitcoin blockchain is used as a distributed, write-once database to store the registry records. Combined with cryptography, this makes the registry global, robust, and tamper resistant. Using cryptographic hashes has the additional benefit of shielding the personal identity of the parties, yet enabling identity to be proved if needed.
The registry tracks the history of ownership, which is what the art world calls “provenance”. They also deal with the concept of “limited editions”, with N > 1 identifiable legal copies. (Note that “Provenance” is a much more general concept, and something like Ascribe could be very useful for, say, recording the processing history of a scientific dataset.)
This is basically the first implementation.
The business model is pretty simple. There are five things you can do:
Basically, the registry is supported by ‘taxing’ transfers (sales).
This first version lets me find an authorized copy of a digital object, or obtain rights to it. What it doesn’t do is prevent or even detect unauthorized copies or uses. The latter will require large scale search and comparison, to detect objects that are “similar enough” to be considered copies, and compare them to the registry.
Overall, the system is interesting and, with the “easy legals”, may well be useful.
There are a couple of important lessons from this. First, the blockchain is useful but scarcely sufficient to implement intelligent services. And second, “smart contracts” on the blockchain require “the legals”.
However decentralized and global the Bitcoin blockchain, the records entered by Ascribe are interpreted under German law. Ascribe aims to provide legally useful records, but the contracts are up to the parties involved.
- McConaghy, Trent and David Holtzman, Towards An Ownership Layer for the Internet. ascribe GmbH White Paper, 2015. https://firstname.lastname@example.org/ascribe whitepaper 20150624/digitalwork/ascribe whitepaper 20150624.pdf
- McGrath, Robert E., Introductions to Making at a Community Fab Lab: Experience and Perspectives. Champaign Urbana Community Fab Lab White Paper, Urbana, 2013. https://robertmcgrath.files.wordpress.com/2013/05/white-paper-2013-mcgrath-v11.pdf