Open Index Protocol

From OIP Wiki
Jump to: navigation, search
Open Index Protocol
An 'open jungle' blockchain specification for a worldwide database
Initially Developed By Blockchain Technology Group LLC
Live Demo oip.io

The Open Index Protocol is a specification for decentralized publishing, distribution and payments of any digital media. The specification includes:

  • How a proof-of-work blockchain[1] and interoperable transport protocols create a shared data layer for digital content
  • How metadata, terms of use, file addressing, and payment information are stored in the blockchain
  • How to search and browse the shared data layer
  • How the blockchain security incentive is connected to the value of the content published to the shared data layer
  • How file persistence & legal compliance are ensured
oip simplified stack
comparison of web vs oip stack

Open Index Protocol uses the Flo blockchain for the meta-data index and interoperable transport protocols for file storage/distribution and payments. Current implementation supports IPFS for file storage/distribution and Bitcoin, Litecoin and Florincoin tokens for payments; support for additional file and value transport protocols is on the roadmap.

History

In February 2014 Devon & Amy James proposed decentralized applications ArchiveChain[2] and MovieCoin[3] on the Ethereum community forum. Early development proved a shared data layer for any digital content is more efficient than separate appcoins for each, so these initial ideas were combined to become Open Index Protocol. Significant contributions to the specification have been made by Ryan Jordan, Ryan Taylor, Skylar Young, Jeremiah Buddenhagen and lead Florincoin developer Joseph Fiscella.

The initial proof of concept archived social media data into a blockchain and visualized the data as an interactive word-cloud. It was demonstrated in October 2014 at Inside Bitcoins, Las Vegas[4]. Next, streaming audio and video were added using the BitTorrent network. On Feb 19, 2015 the first piece of media was published and retrieved using a 100% decentralized system. Bitcoin has been used for paid content since the initial public release in April 2015 [5]. In May 2015, the primary file storage/distribution network was changed to IPFS and support for BitTorrent was deprecated. At each stage, a user-facing browser or visualizer was built alongside specification development [6]. Initially the browser was known as ‘The Decentralized Library of Alexandria’ and the specification as the ‘Alexandria protocol.’ In a demonstration at the Decentralized Web Summit hosted by the Internet Archive in June 2016, Sir Tim Berners-Lee said the names were confusing and suggested they be changed [7]. The browser name was changed to ‘Alexandria’ and the specification name changed to ‘Open Index Protocol.’

Problem

Central points of failure plague current digital distribution architecture (Figure 1). The system is vulnerable to attack demonstrated by problems like hacked personal information[8], spying[9][10], deplatforming, demonetization and censorship[11], as well as direct technical attacks on network infrastructure itself[12]. Current architecture also suffers from wasteful redundancy[13] which increases operational costs[14]. Network speed is fragile because speed and popularity are negatively correlated (Figure 2).
A sketch of network shapes from On Distributed Communications.
Figure 1
A sketch of peer download speeds comparing HTTP to p2p networks
Figure 2

Solution

Zero central points of failure. Every component of Open Index Protocol is distributed. The specification is a permissionless system[15] that creates an ‘open jungle’ where anyone can publish, distribute and sell. The Open Index Protocol changes the fundamental economics of content distribution by transforming individually-negotiated, contract-based services into digital commodities that can be exchanged with fungible tokens. The system expands the sharing economy[16] to new services including front end platforms, social media influencers, proof-of-work miners and file storage/distribution providers.

Why Now?

Twenty years ago, print media changed forever when the ‘open jungle’ of the internet democratized information exchange[17]. Walled garden services[18] like America Online, CompuServe & Prodigy did not initially use HTTP, but eventually adopted it because the open jungle of the World Wide Web was growing in popularity[19]. Today, walled garden content distribution services like YouTube[14][20][21], Spotify[22][23][24] & Netflix[25][26][27] are ripe for similar disruption by the open jungle of Open Index Protocol.

The current digital content distribution industry is in crisis. Artists blame confusing contract terms. Audiences are so frustrated by the difficulty of accessing content they resort to piracy[28][29]. The industry as a whole is struggling.

The root of these problems is twofold: 1) current decentralized hub and spoke distribution architecture cannot support market demand as efficiently as a distributed network, and 2) current solutions were constructed when the technology required for a decentralized and open system did not yet exist.

Since 1999, p2p file sharing[30][31], blockchain[32] and other decentralized technologies have been invented to solve these problems. Open Index Protocol defines a specification for how to use these technologies to create a permissionless system for decentralized publishing, distribution, access and payments of any digital media.

Features

  • Open - Permissionless system for publishing, distribution & payments of any digital media. The shared data layer functions as a public open index; blockchain protects metadata, peer-to-peer file storage/distribution ensures access, and digital currency streamlines payments. Transport layers for file storage/distribution and payments are fully interoperable.
  • Efficient - Distributed networking reduces system overhead, cryptographic tokens decrease administrative management of content & payments, p2p file storage/distribution increases playback performance and user payment process is simplified.
  • Independent - Unique Terms of Use controlled by individual user, increased revenue channels for publishers (including direct payments in social media).
  • Anticensorship - Open Index Protocol is fully transparent. Full global state replication of index data, transparency of all publish attempts including fails, and human readable index data ensure content data is transparent and auditable. Prevents deplatforming, demonetization and censorship.
  • Secure - Index is protected by proven proof-of-work blockchain and contributing to security is incentivized. The commercial value of content is connected to and drives the security incentive. System enables collective defense against attack, ensures liquidity of tokens for publishers, and automates reliable profit margins for miners and traders.
  • Antifragile - System has positive sensitivity to increases in volatility. Free market economics strengthened by interdependent incentive structure; system fees drive network security. Chaotic market inputs cause convex response, improving network incentive for all participants (publishers, platforms, index mining and storage/distribution of files). Value capture is split between subjective/objective services which benefits overall network growth. Interoperable transport layers can adapt to changes in market preferences.

Design Philosophy

The system design objective is to serve the digital media distribution market with an open specification using decentralized technology. To meet the needs of this large and diverse industry, the specification is:

  • An "open jungle"[33]
  • As interoperable as possible with as few rules as possible
  • Capable of serving all distribution and monetization models: free, ad based, pay per view, subscription, metered, non-financial exchange (seeding), rental, lifetime access, stream, download.

Example uses: backend shared data layer for blog entries, Wikipedia articles, Netflix shows, YouTube videos, Spotify songs, SoundCloud remixes.

Assumptions

  • A sufficiently diverse marketplace will include rational actors acting in their own best interest who understand it will benefit them to use standards & obey the law
  • A proof-of-work blockchain-based shared data layer is the most efficient technology available to protect information freedom and resist entropy of information access
  • Incentive structures influence outcomes. A trust-minimized, permissionless system with a sustainable market-based incentive structure fosters cooperation, competition, and iteration; continuously improving product-market-fit
  • At present, some human governance is necessary. Open Index Protocol Working Group is a potential central point of failure, vulnerable to social attack. This risk is minimized with a transparent governance structure, weighing feedback from all system participants: publishers, platforms, influencers and miners

Principles

To ensure system integrity, the specification was built with two inviolable principles: permissionless system design and limited standardization

Permissionless system design

All networks used by Open Index Protocol are permissionless, anyone can join the networks without requiring authorization from another party. Examples of other permissionless networks include BitTorrent, Bitcoin, IPFS and Ethereum. Permissioned layers can be built on top of a permissionless system for use-cases where identity and permissioning are appropriate, however the opposite is not possible.

Open Index Protocol is a permissionless system because freedom of information requires a free and open shared data layer for digital media to ensure media persistence, censorship resistance, security, and sustainability.

Media persistence and censorship resistance are the foundational values behind the permissionless system design. Index and file persistence are secured via: decentralized networks, incentive alignment with positive sensitivity to market chaos, and transparent governance. Censorship resistance is protected via: full global state replication of index, human readable data, auditable records (all publish attempts including fails), and transparent immutable ledger.

Security vs convenience friction is reduced through permissionless system design. API endpoints are universal; they can be looked up through a localhost or through any protocol compliant host. When security is prioritized over convenience, the end user or service provider runs a local protocol daemon and full node. When convenience is prioritized over security, the end user or service provider accesses the system through hosted full nodes. Both local and hosted addresses have the same url format, making host addresses interchangeable.

Sustainability of the system is balanced and maintained with interdependent incentive alignment; the commercial value of the media is connected to and drives the system security incentives.

Limited Standardization

Open Index Protocol specification limits standardization such that the incentive structure benefits assimilation, and does not threaten negative restriction. A salutary approach is used; the specification defines standard use of objective data (index, file addressing, payments), it does not standardize subjective, service-based uses of the data (proprietary discovery algorithms, user experience, user interface). The co-opetitive model benefits from cooperation on objective data, competition of subjective services, and collaboration for iteration; system incentive design improves potential for product-market-fit.

Standardization is powerful; Open Index Protocol benefits from all three fundamental characteristics of standardization:

An example of the power of standards is the rate of google searches compared to emails sent. Google is the most popular individual website in the world, but it’s index and search algorithm are proprietary, whereas email is a standard; for each google search[35] performed, about 40 emails[36] are sent.

Economic Principles

Blockchain changed how value flows through systems and how value is captured.

Joel Monegro coined the term ‘Fat Protocols’ in his landmark article, where he compared the opportunity for value capture on the Web with the opportunity for value capture with Bitcoin. James Kilroe added to this framework with his analysis of value flow and opportunity for value capture with a model he calls ‘Application Protocols.’

Jake Brukhman brought nuance to the conversation with his criticism that the divide between application layer and protocol layer can be drawn at any level in the stack. Although Brukhmans criticism is valid, for the purposes of evaluation, we believe the terms ‘application layer’ and ‘protocol layer’ have been accepted as part of the common lexicon for these ideas and are useful for this discussion. We agree with Brukhman that the idea of value pooling in base protocols is a misapplication of Monegro’s modeling of value networks and that what is important is how the value flows through the layers of the functionality stack.

We propose an additional value capture model called ‘Salutary Protocols,’ wherein the boundary between application and protocol layers is determined by the divide between subjective and objective kinds of work - the application layer performs subjective work, the protocol layer performs objective work and both capture value based on how competitive they are. In a Salutary Protocol system, the protocol layer drives some of the value that it facilitates to the application layer, creating financial incentive for developing competing applications.

Salutary Protocols add value to the application layer while supporting the network effects of the protocol layer. This is possible because blockchains are effective at capturing value at the protocol layer. Salutary protocols are sustainable because it is in the best interest of all participants to contribute to updating, securing, and expanding the infrastructure of the protocol layer. In this discussion, we review the basic frameworks discussed by Monegro, Kilroe, Brukhman and others, and analyse the new value capture model we call Salutary Protocols.



Thin Protocols

A sketch of thin vs fat protcols.
Joel Monegro's Fat Protocol's.

TCP/IP and HTTP/World Wide Web are thin protocols because all of the value (often in the form of data - ie Google, Facebook, but in some cases as a percentage of financial transactions - ie PayPal, Netflix, Amazon) is captured at the application layer and none is captured at the protocol layer.

In this model, application layer products and services must have a closed, proprietary system and maintain end-to-end control to enforce value capture and terms of use. This distribution network topography requires that each company maintain data centers located within close proximity of their end users in order to provide enough redundancy of files for them to be readily available to users. The network speed is fragile because data transfer speeds and popularity are negatively correlated, the inefficiency of this system can result in slow playback and poor quality for end users and increased overhead for service providers.

The World Wide Web is comprised of walled-garden applications because of its thin protocol model. Although the Web offers no method for value capture at the protocol layer, it offers enough standardization and capability to have strong network effects.The World Wide Web is the best example of the importance of competition at the application layer -- one protocol with more than a billion applications on it.[37]


Fat Protocols

Bitcoin is an example of a fat protocol because the majority of value capture happens in the protocol layer by miners, and the application layer can only capture value by offering additional services beyond the protocol functionality. Fat protocols offer no direct incentive for the developer community to create application layer products and services, thus options at the application layer may lack of competitive options for users

Fat protocols threaten to perpetuate the walled-garden problem and may make it worse - the potential for severe long-term consolidation at the application layer in the fat protocol model, could in turn compromise the integrity of the underlying protocol. The economics of the system are fragile.


Further, commoditization will drive the price down over time. Blockchain protocol tokens can be used to buy fungible commodities for Storage/distro as a Service, Processing as a Service and Security as a Service. The combined properties of forking and interoperability will lead to the commodification of these services, which in turn will drive the price of these down over the long term and could substantially change the market dynamics.



Finally, projects that have raised money to fund development with the Fat Protocols model are vulnerable to instability because they are threatened by short convexity. In her article ‘Short Convexity,’[38] Jill Carlson explains the system design and economic incentive structures that result in short convexity and the potential ramifications for this method of fundraising. Carlson explains that the potential problem of fundraising in a Fat Protocol model is that the company is holding the tokens they are dependant upon for cash flow, so if they need to sell tokens when the market is low they are faced with the double edged sword of having to sell at a low and may drive the price down further by selling the tokens.


Application Protocols

‘Application protocols’ apply a set of standards to a use case and have tokens that are unique to the Application Protocol, but they also use base layer protocols with tokens for various functions. James Kilroe used aggregation theory in his analysis of Application Protocols to show how value will move toward the most interoperable systems because integrating multiple verticles will increase overall value capture.

Kilroe identifies the primary connection for the end user is to the platform they use. The Application Protocol model enforces this connection by requiring end users to use the Application Protocol token, similar to the way one must use tickets at an amusement park. The incentive design in this model is that end users will share in the value captured by the economy through owning the tokens and the assumption is that this will incentivize them to share and grow the system.

Although this incentive model may work short term, it is not an effective long term strategy. At scale, end users will not care which tokens they hold. End users will hold whichever reserve currency they prefer and will pay for services that require unique tokens by automatically swapping their reserve currency tokens for whatever is needed for the transaction. The market will move toward systems that offer the most interoperability and specifications like Lightning Network and Cross Chain Atomic Trades will make trades immediate. The inherent weakness in the economic incentive design of this model is similar to the ‘Fat Protocols’ model because it creates fragility of the Application Protocol token price.

Further, this approach replicates the current problems of the walled-garden model. We agree with Kilroe that platforms should serve end users, however and essential component of the crypto-economic design at scale is missing -- the incentive for competitive user-facing applications.

Instead of the platform enforcing its connection with the user via required tokens, the platform should get market-based feedback from its users. The necessity of competition at the application layer has often been overlooked in these discussions, for example Kilroe calls applications “glorified UX.” Protocols that have only one application built upon them are weaker and more fragile than protocols that have many, and incentive for competition at the application layer is the key differentiating factor.


Salutary Protocols

A sketch of a salutary protcol.
OIP's Salutary Protocol.

We propose a Salutary Protocol model in which the specification creates opportunity for value capture at both the application and protocol layers.

Salutary protocol systems maximize efficiency by separating subjective and objective work and empowering the marketplace of users to define their own unique combinations of services and pricing.

Service providers of subjective kinds of work, like user experience, content discovery, and filtering lists capture the value assigned to the application layer by users. Application layer services compete based on features that are important to their users. The marketplace can serve the full range of user demand with a spectrum of options for products, services, price points, interface preferences, quality compared with price and transaction confirmation duration, etc.

Objective work like index security, transaction confirmation, and file storage capture value at the protocol layer via blockchain tokens. The tokens function as fungible commodities that represent these services. As Kilroe & others have suggested, it is likely that over time commodification will drive down the market price of these services.



Interoperable base layer protocols will function as fungible commodities and be used for objective work; the market for these services will flatten and compete toward maximum efficiency.

Applications compete based on subjective qualities that matter to end users such as user interface, content discovery and filtering.

Salutary protocols leverage the benefits of interoperable base layer protocols and create financial incentive for application interfaces to compete for end users.


Fees examples

Thin protocols do not capture value. Without the ability to collect fees at the protocol layer, thin protocols have no direct financial incentive. Organizations like the W3C that are responsible for maintenance and future development of thin protocols are limited to non-profit funding models like member dues, research grants, private sponsorship and donations.[39] Products and services built on thin protocols can only capture value at the application layer and their value capture methods have many problems -- broadly, limited transparency and subvert data theft.

Example A: When a video is viewed on YouTube, the base level protocols that make it possible do not capture value (TCP/IP, HTTP, YouTube’s internal proprietary protocols). Youtube’s application layer services maintain end to end control of the content and captures all potential value in the form of advertising revenue and user data. The application layer must add value beyond the protocol function.

Example B: When a party to party payment is facilitated by PayPal, the base level protocols that make it possible do not capture value, they only host the interface with which the parties interact. The transfer of value is controlled by a combination of proprietary user accounts, network effects between users who need the service & paypals relationships with banks. As such, the protocols involved capture no value, PayPal is the only party which can capture any of it, and it splits some of that value with its banking partners.

Fat Protocols value capture opportunity is almost entirely at the protocol layer. In this model the protocol layer captures most of the value. In the Bitcoin network the majority of value is captured by miners in the form of tx-fees and the temporary subsidy of block-rewards. Application layer value capture in the Fat Protocol model can only occur with value-add services beyond the function of the protocol. Application layer products and services are competing toward zero margins because the protocol layer can serve its primary function without the optional value-ad services offered by the application layer. Thus, opportunity for value capture from application layer fees is very small because of the potential for competing services that cost less or are free.

Example 1: When a party (A) to party (B) payment is facilitated by Coinbase, the Bitcoin protocol is what is facilitating the value transfer. When it happens, a protocol worker (a POW miner) is the only party that can capture a portion of the value, in the form of a tx-fee which could range from fractions of a penny to a few cents (to a few dollars in uncommon circumstances). Coinbase has no opportunity to capture any of this value.

Party A sends $100 in BTC to Party B.

Protocol worker C captures ~$0.01 from it, and ~$99.99 is delivered to Party B.

Example 2: However, if the sending party doesn’t yet have Bitcoin, Coinbase can offer the service of exchanging fiat money from their bank account to Bitcoin. Because this requires banking relationships and regulation compliance they have some exclusivity over the service, which lets them charge a fee similar to the fee that most merchants are charged for processing credit or debit cards, 1.5%.

Party A pays $101.50 on their debit card to send $100 in BTC from Coinbase to send to Party B.

Coinbase captures $1.50 for the conversion, Protocol worker C captures $0.01, and $99.99 is delivered to Party B.

Application Protocols have the same fundamental economic design as fat protocols. They don’t have workers which capture a portion of the value they are facilitating - they facilitate one service, workers are paid 100% of fees to provide it, but the fees must be paid IN the token. As such application protocols don’t allow the application layer to capture value by providing a service. The only way for a person to “invest” in an application protocol, or earn any revenue with it, is to speculate on the underlying token itself. The weakness of this approach is that its based on the underlying assumption that end users will buy and hold tokens based on the services they can pay for with those tokens. In reality, however, users will gravitate toward holding one or few “reserve” tokens, and then use live-exchange services to buy the tokens they need for various applications at the time they want the service. This completely undermines the value proposition of application protocols.

Theory relies on users holding Application Protocol tokens to use the application. This will hold true for only a limited period of time until the reserve token takes over, and then token value will depend entirely on velocity. There is not a standard

Salutary Protocols capture value at both the application and protocol layer. In this model, value is captured for subjective work at the application layer, and objective work at the protocol layer. The application layer fees are part of the total transaction amount.

Example: Party A sends some value to Party B, Party C is an optional participant offering some subjective service to make the transfer go smoother, and Party D is at the protocol-layer doing the objective work that A, B and C all depend on.

Party A spends some amount (consumer of content or coffee)

Party B gets the lion’s share of it (creator of content or coffee)

Party C gets a fee if it helped move things along efficiently (content platform or LN channel)

Party D gets a fee for ensuring the whole system is running (bitcoin miner or bitcoin miner)


How OIP uses the salutary model

OIP creates competition at the application layer for platforms and influencers and at the protocol layer for miners and other service providers.

Key
Application
Protocol
Either/Neutral

The specific steps of OIP’s salutary model

  • Publisher creates commercial record & pays system fee
  • Index miner wins block with record and receives publisher’s fee in addition to block-reward
  • Platform hosts & displays record
  • Influencers promote record via social media
  • End user discovers record promoted by influencer and purchases it thru platform
  • Payment received from end user is automatically split to publisher, platform and influencer based on split amounts listed in record


When a publisher creates an record in OIP, they state the amount of value capture opportunity available to parties who distribute or promote their content. The value assigned by the publisher of the record is then captured by application layer product and service providers like platforms and influencers. Application layer revenue is performance based. The opportunity for value capture in the system is transparent and explicit, and the method to participate increases interoperability and reduces backend work.

FLO blockchain miners and file storage/distro miners capture value at the protocol layer. In addition to the block-reward, FLO miners capture value from fees paid by publishers. All records published to the system require a fee. Free records require only the cost of putting the data into the blockchain which is so low it could be considered trivial. The fee for commercial records is derived from the commercial value of the content itself. These fees balance the relationship between the security of the system with the value of the records in the system. The effect is similar to the sink (fee) advocated by Vitalik Buterin because the fees are transparent and explicit and increase the stability of the system.[40]

The interdependent structure of OIP’s salutary model has market pressures from all stakeholders. Alignment of incentives through an open market with closed loop fees is the key to the system’s potential for antifragile growth and long term sustainability.

Example using OIP:

Party A publishes some valuable content using Open Index Protocol. Because there is a $1 price tag on the piece of content, Party A spends $1 in FLO tokens as the publish fee.

Party B (protocol layer) is the FLO blockchain miner which verifies the block with this piece of content in it.

Party C is an end user who wishes to pay Party A $1 to watch the content.

Party D, a platform (application layer), hosts the application which Party B used to enjoy the content.

Party E, a bitcoin miner (protocol layer), verifies the transaction sent to Party A.

In this example, this is how the splits will result:

Party A will spend a 1-time fee of $1 in FLO to share the piece of content, and Party B will capture it.

Party C will spend $1 to enjoy the content, of which, Party D will capture $0.20, Party E will capture about ~$0.01 of it, and Party A will receive the remainder of the $1, about $0.79.

How Lightning Network uses the salutary model

Lightning network creates incentive for competition at the application layer between channel operators. Channel operators will compete to serve different parts of the market by offering a variety of options like settlement reliability, blockchain settlement frequency, transaction amount size and channel operator fee. Channel operators can also function as market makers and bridge funds if transaction completion is delayed by a node going offline.

Example: User opens a channel with a $100 deposit and uses it for small daily purchases with once per month settlement. User spends a few dollars each day for coffee and spends the full amount over the course of the month. Based on the terms of the channel, a small percent will go to the channel relays as fees and the majority will go to the coffee shop. The total amount spent is $100.

Party A is end user with a budget of $100 per month for coffee

Party B is his favorite coffee shop

Party C are the relay nodes in the LN channels he uses

Party A gets coffee, Party B gets the lion’s share of the money, ~$99, and Party C gets a tiny sliver of the money for facilitating the transactions, ~$1.


Comparison of digital distribution services

Service Offering Cost to Users Revenue Creator Share of Revenue
Open Index Protocol Self Publishing of Downloadable and Streaming Music, TV, Film and Videos A La Carte, Subscription and Ad-Based Varies based on use Completely up to creator As much as 100% to creator
iTunes Purchasable Music A La Carte Only $0.99-$2.19 per track Varies based on use 70% to creator
Spotify Streaming Music Subscription & Ad-Based $9.99 per month ~$0.0072 per play 70% to creator
Pandora Algorithm-Based Streaming Music Subscription & Ad-Based $4.99 per month ~$0.0014 per play 70% to creator
SoundCloud Self Publishing of Streaming Music Subscription & Ad-Based $7-$15 per month Unknown Unknown
Netflix Streaming TV/Film Subscription Only $7.99 per month Unknown Unknown
YouTube Self Publishing of Streaming Video Subscription & Ad-Based $9.99 per month ~$0.0005 per play 55% to creator
Crackle Streaming TV/Film Ad-Based Only $0 Unknown Unknown

See also

  • Alice and Bob: An explanation of the various kinds of users and service providers in the Open Index Protocol eco-system.
  • Specification: The rules, variables, formulae and thresholds that make up Open Index Protocol.
  • Message Protocol: The JSON schema of various kinds of OIP messages and definitions of all terms used within them.
  • Security Considerations: the questions, solutions & security considerations that have been analyzed.
  • Back to the Main Page

Other Open Index Protocol Applications in Development

  • Token.fm streaming music player
  • Caltech/Jensen Lab's Tomography Public Database
  • Robo3D's "3DRM" marketplace
  • YouTubexit.com video archiver
  • eVue Digital Labs

References

  1. https://en.wikipedia.org/wiki/Proof-of-work_system
  2. https://forum.ethereum.org/discussion/440/
  3. https://forum.ethereum.org/discussion/427/
  4. https://letstalkbitcoin.com/blog/post/lets-talk-bitcoin-154-on-the-road-again
  5. Alexandria v0.4.0 Video Demo
  6. Alexandria v0.5.1 Video Demo
  7. Presentation at the Decentralized Web Summit
  8. Washington Post: Data of 143 million Americans exposed in hack of credit reporting agency Equifax
  9. 9 Examples of How Smartphones Are Eavesdropping on You
  10. Wired: A Hacker Can Turn an Amazon Echo into a Wiretap
  11. BBC: Turkey social media ban raises censorship fears
  12. https://www.theguardian.com/technology/2016/oct/26/ddos-attack-dyn-mirai-botnet
  13. Juan Benet IPFS Presentation
  14. 14.0 14.1 YouTube: 1 Billion Viewers, No Profit; The Wall St Journal
  15. Blockchain – What is Permissioned vs Permissionless?
  16. Sharing Economy;wikipedia.org
  17. http://www.businessinsider.com/chart-of-the-day-craigslist-vs-newspaper-2009-6
  18. https://blog.codinghorror.com/avoiding-walled-gardens-on-the-internet/
  19. https://www.internetworldstats.com/emarketing.htm
  20. http://nymag.com/selectall/2017/12/can-youtube-survive-the-adpocalypse.html
  21. https://www.theguardian.com/technology/2014/feb/24/zoe-keating-itunes-spotify-youtube-payouts
  22. http://www.abc.net.au/news/2017-09-06/digital-music-streaming-rising-but-spotify-losing-money/8875188
  23. https://www.digitalmusicnews.com/2015/06/11/the-more-money-spotify-makes-the-less-artists-get-paid/
  24. https://www.bloomberg.com/news/articles/2017-06-15/spotify-losses-widen-after-music-site-flags-accounting-errors
  25. https://www.fool.com/investing/2017/11/09/media-mergers-are-bad-news-for-netflix.aspx
  26. https://money.usnews.com/investing/stock-market-news/articles/2017-12-11/walt-disney-co-dis-twenty-first-century-fox-foxa-netflix-inc-nflx
  27. http://kicks99.com/2017/10/09/bad-news-netflix-subscribers/
  28. https://www.smh.com.au/technology/four-myths-about-online-piracy-whos-doing-it-and-why-20150309-13z2vm.html
  29. http://theoatmeal.com/comics/game_of_thrones
  30. http://www.bittorrent.org/beps/bep_0003.html
  31. https://ipfs.io/ipfs/QmR7GSQM93Cx5eAg6a6yRzNde1FQv7uL6X1o4k7zrJa3LX/ipfs.draft3.pdf
  32. https://bitcoin.org/bitcoin.pdf
  33. https://www.wired.com/2017/04/tim-berners-lee-inventor-web-plots-radical-overhaul-creation/
  34. What does “permissionless” mean?
  35. http://www.internetlivestats.com/one-second/#google-band
  36. http://www.internetlivestats.com/one-second/#email-band
  37. http://www.internetlivestats.com/total-number-of-websites/
  38. https://medium.com/@jillcarlson/short-convexity-8793f18629bb
  39. https://www.w3.org/Consortium/facts.html
  40. https://vitalik.ca/general/2017/10/17/moe.html