"Why Fake News Is Relevant" - Introduction to the Userfeeds Protocol
3 likes6,625 views
This document introduces the Userfeeds Protocol, an open protocol for establishing information relevance in crypto-economic networks. It discusses the problems with how information relevance is currently solved by centralized platforms and proposes using a decentralized protocol and blockchain to aggregate evaluations across contexts to extract both vertical and horizontal relevance. This would allow tokenization of relevance domains and open up new opportunities for consensus-based and personalized relevance. The current implementation includes a Userfeeds Relevance Platform built on Ethereum with ranking algorithms and interfaces. Several potential use cases are described, including a Link Exchange platform to monetize token-based communities and applications.
"Why Fake News Is Relevant" - Introduction to the Userfeeds Protocol
- 1. Introducing the Userfeeds Protocol An open protocol for establishing information relevance in crypto-economic networks. Why “Fake News” Is Relevant @maciejolpinski
- 2. Research Development Protocol + Software platform development: Userfeeds Relevance Platform LinkExchange.io Economics of attention & information relevance: Userfeeds Protocol What We Do?
- 3. What happened in 2017 Seed Funding Initial Idea + Prototype Platform + Apps Live Userfeeds Relevance Platform Link Exchange Team Ramped Up to 7 People (5 Software Engineers + 2 Founders) January 2017 January 2018
- 4. It didn’t solve the problem of information relevance. HTTP/TCP/IP/SMTP solved the problem of information transmission. Relevance is solved by the application layer companies today. The Problem How to solve the relevance problem for the Web 3.0 stack?
- 5. What is relevant information?
- 6. Naftali Tishby Fernando C. Pereira William Bialek The Information Bottleneck method https://arxiv.org/abs/physics/0004057
- 7. “We define the relevant information in a signal x ∈ X as being the information that this signal provides about another signal y ∈ Y .” Source: The Information Bottleneck method paper https://arxiv.org/abs/physics/0004057
- 8. “Examples include the information that face images provide about the names of the people portrayed, or the information that speech sounds provide about the words spoken.” Source: The Information Bottleneck method paper https://arxiv.org/abs/physics/0004057 Mike
- 9. Relevant information in behaviours (x) of people on the Web (X) is the information that it provides about other behaviours (y) of other people on the Web. (Y) What if we could extrapolate this definition to the broader definition of “relevance” on the Web:
- 10. Increasing relevance on the Internet is all about pursuing the ability to elicit “predictable” behaviours in others. “Fake news”, digital addiction are natural consequences of companies seeking more relevance / predictability.
- 11. Single “Evaluation” is an Atomic Signal of Relevance in Networks Website Website Person Photo Bank Account Purchase Like Link “Anchor Text” Product
- 12. Aggregated evaluations can be “compressed” by algorithms to extract relevance (predictability) Platform Google Facebook Amazon Context Keyword Social Circle Consumer
- 13. Vertical Relevance Subjective Context Dependent Individual Feeds ‘Silicon Valley’ business model Personalization Algorithms In Space Likes, Links, Followers
- 14. selling advertising product The aggregator business model of the Web 1.0 and Web 2.0 Extract relevance from as many contexts as possible and convert to money buying advertising product $ $ $ $ $ $ $ $ $ $ $ $cash stock price
- 15. Horizontal Relevance Objective Context Independent Shared Ledgers ‘Wall Street’ business model Consensus Algorithms In Time Money, Assets
- 16. $ $ $ $ $ $ $ $ $ $ $ loss of information during transitions Silicon Valley Data Silo Wall Street Data Silo You are here You are here
- 17. Internet 3.0 will create an explosion of horizontal relevance domains via tokenization. Each of them will be individually evaluated by the market. This will open up a new, temporal axis for establishing future relevance of ideas and information. $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $
- 18. Horizontal Relevance Vertical Relevance Token Token Price Token Token Price Token Token Price Tokenisation ? Question: “how to go back from consensus to personalization?”
- 19. “Proof of evaluation” is a basic signal of relevance. One Message Type for Both Vertical and Horizontal Relevance A Bmetadata C Cryptographic Key Arbitrary StringArbitrary Metadata Userfeeds protocol
- 20. One message type for all evaluations. Like Photo Link URL “anchor text” Token Transfer BlockchainAddress Cryptographic Key
- 21. Ranking API Application Ranking AlgorithmRanking Algorithm IPFSEthereum Blockchain Amazon S3 Evaluation Display Rankings Aggregation Publication Protocol Layers
- 22. Aggregated Graph of ‘Proofs of Evaluation’ ImageTextVideo
- 23. Infinite Evaluation Cycle Relevance Layer as Integral Part of the Web 3.0 stack Ranking API Ranking API Observer Application Observer Ethereum Blockchain IPFS Amazon S3 Amazon S3 Ranking Algorithm Ranking Algorithm Ranking Algorithm Ranking Algorithm IPFS Ethereum Blockchain Amazon S3 Application
- 24. Data layer Processing layer Display layer Current Implementation of The Protocol Userfeeds Relevance Platform Ethereum Blockchain X Database AlgorithmsAlgorithms API InterfaceInterfaceInterface Userfeeds Relevance Platform Blockchain Y
- 25. Today Proprietary and Separated Evaluation Graphs In Space In Time
- 26. InSpace In Time Userfeeds Protocol Open and Integrated Evaluation Graphs
- 27. Personalization Consensus Userfeeds Protocol Relevance That Combines Consensus With Personalization
- 28. Use Cases
- 29. It’s like Google AdSense for Token-based Communities Link Exchange is Coming Soon Click on an ETH icon to see an early preview https://userfeeds.io 18 3.9212 Boost 5 In slot 15% Pay by Metamask33.0001 E T H 0% 100% 72.3% Available: 15.032 ETH LinkExchange.io Token holders Publishers Advertisers
- 30. Link Exchange for Personal ERC20 Tokens
- 33. … and it works with Augmented / Mixed Reality too
- 34. cryptopurr.co “Social Network” for Cats … or any ERC721 token
- 37. State Of The Dapps & LinkExchange - coming soon Promote dApps with Ether