I've been thinking some more about the infrastructure to make this concept work.
Let's suppose that the copyright offices are willing to consider compulsory or statutory licensing, i.e. if someone wants to distribute copies of a work, they can do that, but they have to pay a fixed licensing fee per copy distributed. These fees are usually collected by collection societies
which are set up to monitor distribution, collect royalties, and remunirate creators (pay them the royalties).
Now imagine a file sharing protocol (not necessarily peer-to-peer) that accepts uploads from anyone, but requires certain information to be entered to upload the file, namely the creator of the content. (It would be helpful if content could be checked against existing content held in the system for matches, as well.) The filesharing protocol notifies the collection society whenever the file is downloaded using the protocol, and as part of the download, the collection society assesses the statutory licensing fee.
The protocol could support collections of files and/or download links by third parties, so services could develop of finding and recommending specific content, possibly supported by advertising, similar to radio stations.
The protocol could support direct payments to the one who provided the file (the converter). Better copies (e.g. books with formatting and spell checking) might cost more. The identity of the person who provided the file would not be anonymous, but the system would be legal, so this should not present a barrier.
The protocol could support payments by the server rather than the downloader, so services could create alternate product packages, e.g. time-based subscription models, or could choose to derive more of their revenue from advertising and lower the portion of the statutory licensing fee passed on to customers.
The protocol could support some kind of affiliate program, so that someone who successfully recommends content could be compensated for their effort in locating and selecting/reviewing content.
A peer-to-peer version of such a protocol could support a fee for serving the file, rewarding a distributed service model, and further encouraging people to be honest about the source of a file, hopefully ensuring that the actual creator of the work is compensated.
Fees for works by creators unknown to the system would still be assessed for a period of time (e.g. 3-5 years), after which if no creator had been identified the work would be designated "orphaned" and would no longer bear a statutory license fee.
- Someone could upload a file and lie (or make an error) about who created it, causing remuneration to the wrong party if the file is downloaded. However, a non-anonymous system as described in the options above could help limit this problem.
- Content creators/owners may object to such an uncontrolled system of distributing content. (Read Cory Doctorow's essays for examples of how vaudeville performers objected to the first radio broadcasts of musical performances.)
Key assumption: the easier it is to find and obtain content legally, the less likely someone is to bother stealing it. If the legal protocol/network has all the content people want, is easy to find, is safe, and is reasonably priced, far more customers will use it than will use the darknet. Some of the options above actually provide incentives to users to share files legally, as well, which the darknet typically does not.
It turns out, of course, that none of these ideas are new. The Electronic Frontier Foundation
has suggested this idea among others (e.g. sales taxes on peer-to-peer file-sharing software, blank storage media, or even Internet access itself) as a constructive way to deal with filesharing.