Encyclosphere

Encyclosphere Seminar Plan

> Sign up for the seminar here! <

Read more about the planning strategy for the seminar below.

Version 3
February 2021
Larry Sanger

I. Introduction

I want to conduct a combination “master class” seminar and deliberative forum, beginning M 2021. The seminar would cover key questions concerning the design of the encyclosphere.1 It would have two main purposes: (1) to educate participants (as well as anyone interested) about decentralized, open networks such as the encyclosphere will be, and (2) actually to deliberate about the policy issues that will shape how the encyclosphere is organized.

To better understand the seminar’s scope, think about blogs, and particularly about the network or global collection of blogs called “the blogosphere.” The RSS and Atom standards made it possible for blogs to share their content in standardized technical format, which in turn made it possible for blog (more generally, feed) reader software to aggregate blogs and serious edited news articles as well. Naturally, many decisions went into settling the RSS and Atom standards that made the blogosphere possible.

Now imagine that one wanted to build a similar network, but for encyclopedias. The cases are similar, but not exactly the same. The decisions that go into the encyclosphere would be important, maybe especially at a historical moment in which people are concerned both about so-called fake news and about encroaching censorship and centralization. It is very important that we arrive at the best possible set of policies. At the same time, it is important that we have buy-in from developers, entrepreneurs, and network participants. But even those people are not experts about every aspect of the policy that would inform this network.

Therefore, this seminar and forum would accomplish both goals.

II. Seminar Mechanics

Time

The seminar will begin around mid-January 2020 and run at least three months, but possibly as long as nine months, or even longer. It really depends on how long we want to spend studying the material and how long we should discuss the questions. I think it will have to be contingent upon the completely worked-out seminar schedule.

We might take whole weeks off for common holidays (such as Easter/spring break) and for any other major, time-consuming events. But since our guest speakers will need to be scheduled in advance, there cannot be too much “wiggle room” as to the schedule.

Participants

We propose to make the seminar open to the public, but will participation requires a registration process that has a “payment” process. Payment will be optional, but some is encouraged, and if possible we will make it tax-deductible.

Format or “Requirements”

The seminar will combine three basic elements: video (perhaps livestreamed and then saved lectures or interviews), reading, and written discussion. The course would be divided into a number of weekly or biweekly units.

In the video component, our notion is to organize lectures and/or interviews with experts about each unit’s topics, with plenty of time for audience questions after each presentation. Larry would do a very brief introduction to the unit’s issues in his own words, then introduce the guest (or, perhaps, guests), and then the guest would be doing most of the talking, either discussing our focus questions in a lecture, or being interviewed by Larry and then the community. Format (one, the other, or leaving it up to the speaker) is another issue we need to discuss. I am inclined to have two different kinds of videos (about general topics, about specific businesses and tools), and whether there are two videos per week or just one remains to be decided.

The readings would be chosen by Larry and with plenty of input by the community and especially by the speakers. The readings should not be too long—not more than, say, 50 pages per week.

Written discussion would take place in a venue to be determined, but probably in the Encyclosphere.org Forum. Each week would have its own thread. It is in written discussion that much of the actual, detailed deliberation about

I would prefer to have the reading and lecture done at the beginning of the unit, and then we spend several days doing in-depth discussion.

Fundraising aspects

One idea (subject to revision, if someone comes up with a better one) is to register the course through Gumroad.com, which will enable us both to collect money and to charge as little as $0. Presumably, all the materials will be available, but behind some permission that will be given in a mail generated by the Gumroad system. In any case, I think the best way would be to give participant email addresses access to the discussion board and video links.

It might be even better if we just charged directly through the blog. We can probably figure out how to do that quickly enough. That way we don’t have to send people to another website, which could potentially shut us down, and we wouldn’t have to pay any Gumroad service fee (although we might still need to pay a credit card, PayPal, or other service fee for taking on board donations). Pointers and help with this would be most helpful.

Decision-making

The seminar’s purpose is, again, not just to learn how decentralized networks, genuinely leaderless and centerless arrangements of people for special purposes like encyclopedia-publishing and social media, will work and why we should make it. The purpose also includes helping the community arrive at the best possible decisions about what the actual standards will be like.

If indeed we are not just learning, but deciding on the KSF’s official standards for the encyclosphere network, then the KSF Board will need to decide in advance what the procedure will be whereby we decide. My proposal is this: the Board will listen carefully to everyone. We will pay close attention to app developers and publishers who declare that certain situations will make it harder for them to use and support the standards. We will aim to listen closely to and incorporate suggestions from the widest range of possible users and participants.

The decisions we make as we go through the seminar will all be provisional. The final 1.0 version of the standards will be available for an adequately long, well-publicized period of feedback and discussion. Only then will the Board duly publish the standards. We will take our commitments, to our basic principles as well as to the practical needs and requirements of users, very seriously.

III. How to Construct the Seminar

Planning and delivering a new, lengthy, graduate-level seminar (and deliberative forum) using new tools and a new network is not a small task. While Larry could probably execute everything himself, it could be much bigger and better if he has help from the community. We anticipate several main tasks and people needed, as follows.

Syllabus and schedule preparation: This is something Larry will do but in close coordination with whoever wants to help, in the #standards Slack channel. Rough outline below. We can put this up on a Google Docs-style collaborative editing page on Larry’s NAS.

Speaker outreach: We need to start making relationships with guest speakers now. We have agreements or relationships with quite a few different people already, including people already working on the W3C DID protocol, etc., as well as the CEOs of Minds, Everipedia, Ballotpedia, maybe Brave, and various others.

  • Website and software tools preparation: We need to do several things:
  • Launch the website (install Shri’s new design, finalize all the public-facing pages, and move the blog subdomain to the main domain).
  • Choose and install/configure tools for:
    • video (YouTube and Bitchute?)
    • discussion (test the forum; check for password support)
    • payment (WordPress plugin or Gumroad)
  • Design, marketing, publicity: needed for tool support.

IV. Syllabus (Outline of Topics)

Intro: the very idea of decentralizing encyclopedias as an example of how to fix the Internet

  • What went wrong with the Internet? How can we fix it?
  • How do the opponents of Internet freedom think about it?
  • Is Wikipedia decentralized? Why or why not?
  • Why are we thinking about creating an encyclosphere?
  • Looking ahead: Why should developers and publishers adopt uniform, neutral standards for formatting encyclopedia articles (and other categories of content)?

Identity standards

  • What is digital identity?
  • What is digital privacy and why is it important?
  • Does there have to be a digital identity system in order to maintain a useful network of encyclopedic content and ratings? (We do not have one for the blogosphere.)
  • Is it possible to uniquely identify people via any sort of digital identity system? How?
  • What is the best sort of digital identity system?
  • What risks do digital identity systems pose to digital privacy and autonomy? (Think the Chinese social credit system.) Is it possible to guarantee that our digital identity stays in our control, and that it remains private? How?
  • What are the best digital identity systems at present? DID? Are any of them adequate for purposes of the Encyclosphere?
  • Should the Encyclosphere require articles and encyclopedias to be attached to any sort of “digital identity”?
  • What publisher, author, and rater information should be required, if any, as part of the Encyclosphere standards? What should be optional but supported?
  • How do our standards-in-progress measure up against what we have learned in this section? Should we make some changes?

Standards in general

  • How are technical standards best drafted, adopted, and popularized? There seem to be two models for adopting standards, the IETF model and W3C. Which is best?
  • What level of detail does there need to be?
  • Are there dangers associated with standards—and with standards-setting bodies? What can we learn from problems with existing standards and standards-setting bodies?
  • How are standards developed alongside practice (software and data structures in use)? Consider, e.g., the history of the development of the RSS and HTML5 standards. What implication does this have for how we should go about developing Encyclosphere standards?
  • Should we not encourage people to launch projects, or to add features to existing projects, that allow us to determine
  • How do our standards-in-progress measure up against what we have learned in this section? Can we already make some changes?

Decentralized networks in general

  • What is decentralization? What are decentralized networks?
  • Why is the Internet decentralized? How did the decentralization of the Internet come about?
  • Why might one care about this?
  • What are some classic examples of decentralized networks/open protocols? TCP/IP, DNS, email, Blogosphere
  • How well did they work? To the extent that they succeeded, why did they? When they failed, why did they?
  • Is blockchain the technology of decentralization?

Decentralized Distribution, Storage, and Networking

  • BitTorrent
  • IPFS
  • DAT
  • Blockchain
  • LibP2P
  • WebRTC
  • Hypercore

How to decentralize social media

  • What are the special problems associated with the centralization of social media?
  • What

How to decentralize video

 

How to decentralize encyclopedias, part II

 

Encyclopedia standards

  • What should the encyclopedia data format standards look like? What standards exist that would be closest to encyclopedia standards we might adopt? RSS, Atom, ActivityPub…?
  • Should there be a single markup standard for the content of encyclopedia articles? It seems like there will have to be at some point. Is this something readers should do individually (rendering “raw content” from various different formats into the reader’s format), or is it the way that information should be distributed via feeds?
  • And if there is some single content markup standard, what should it be? How fine-grained should it be? How strictly should it be enforced?
    Are there any standards currently in existence that we might use?
  • What should the scope of the standards be? Should we attempt to include only encyclopedia articles, various kinds of reference material, or something broader such as educational content?
  • What sort of metadata should be tracked in the standards?
  • How do our standards-in-progress measure up against what we have learned in this section? Should we make some changes?

Special challenges of the network in general

  • Can the network support a full set of article histories? Should it? How?
  • Should we encourage/support FADs, independent of end-user Encyclosphere reading apps? How important is it to decouple the aggregating of feeds into a database, on the one hand, and building front ends for the database, on the other?
  • How can the network support multiple versions of what began life as the same article? How can thus support content forking, without unnecessarily confusing average end user, who after all wants to read at most only one version of an article, and without unfairly preferring any specific version and allowing exploits to hide versions deemed “bad” by powerful players?
  • Is it technically feasible to allow people to sell content via the network? How? Is this a good idea? It seems important for purposes of getting proprietary encyclopedia publishers involved. But is there even potentially a demand? If this requires encryption, then are there any special technical, legal, or other issues associated with distributing encrypted content over an open network?
  • How, and to what extent, should the network protect the right to free speech? What, if anything, should we do about speech deemed “hate,” yet still legal in some jurisdictions?
  • Should we try to prevent the use of the network to post illegal (copyrighted, libelous, private, secret, etc.) content? We must take into consideration that there are many different laws and jurisdictions.
  • How do we prevent DNS-type attacks, or flooding the network with garbage or spam in attempt to render the network inoperable, while continuing to prevent censorship?
  • What are some unintended consequences of the existence of the network we have in mind to build?
  • How do our standards-in-progress measure up against what we have learned in this section? Should we make some changes?

How might it all go wrong?