1 of 9

A History of Giving

Distributed Reputation for the Sharing Economy

2 of 9

Introduction

  • Sharing economies have worked great for non-rival goods (software, hardware designs, videos, blogs, presentations, forum posts, …)
    • The goods themselves are easily vetted because everybody can have a copy
  • Not so great for rival goods (physical things, in-person teaching). Why?
    • Can't vet items given to others; you don't have a copy!
    • Limited supply: need to decide who does and doesn't get the thing
  • These decisions need a lot of information, and are different for everyone.
  • Systems like StackExchange try to help, but work on only one site.

3 of 9

Motivation

  • I have lots of items at home (old computers, Playstation Portables, cameras, film photographic studio, …) that work, but that I don't need any more
  • Selling to recycle shop feels bad:
    • I don't know where the items go or if they're useful to anybody else: no emotional reward
    • They give me a negligable amount of money, and re-sell at a huge markup, e.g.
      • Books: sell to Book-off for ¥10; they resell for ¥150
      • Sony PSP-1000: S to Comshop for ¥800; they resell for more than ¥3000.
  • But recycle shop does provide some useful services:
    • I don't need to give strangers my address or other personal info
    • Shop vets the item to make sure it works, etc.
    • Shop "finds" (in a way) people who need the item
  • How can I just give to people I don't know well, without selling to recycle shop?

4 of 9

"Real-world" Decisions About Giving

  • When I consider giving an item to someone, I decide based on many factors:
    • Do I know this person? How well?
    • Is he a member of any organizations, groups or movements I'm interested in?
    • What has he done for others in the past? Does he pay things forward?
  • Or I give an item to an organization instead, for them to share or pass on:
    • E.g., give hardware to Tokyo Hackerspace
    • Decisions in a similar way: does the organization do good things?
  • I also have to think about how much work it is to give:
    • Do I need to photograph the item for someone far away?
    • Do I need to pack and ship it?
  • You're can't work from having observed what people did, but you can listen to what people say they or others did.

5 of 9

One Solution

  • Each participant in a giving economy has a ledger recording his participation
    • First entry is a self-signed public key which, standing alone, is anonymous.
    • Linking real-world data (e-mail address, website account, name, location, whatever) makes it pseudonymous or personally identifiable in some way
    • Signing a ledger entry with a private key cryptographically proves which participant made it
  • Participants add entries to their own and others ledgers, e.g.
    • Alice: I sent this item to Bob (single entry on both Alice's and Bob's ledgers)
    • Bob: I received above item from Alice; works great! (on both ledgers)
  • Third parties (trusted or not) can also be involved (more on this later)
  • Note: the items themselves can't be put in the ledger; just like the real world, we're we're working with what people say.
  • Protocol details are in the Design Document

6 of 9

Reputation

  • Once you have someone's full ledger, you can verify it
  • After verification, you can use information from the ledger and other sources to "get to know" the participant, e.g.
    • Is it someone I know? Do they know or have they interacted with other participants I know? Are they associated in some way with organizations or groups I like? What interactions have they had with others? What was the reaction of others to those interactions?
    • Note that, unlike Ebay etc. you have source information; you can find out, e.g., if a friend of a friend was involved in a transaction
  • Everybody evaluates these things differently, and there might be many different kinds of programs to help do evaluations
  • Third-parties can provide verification and evaluation services to others, on- or off-line

7 of 9

Example of Giving With Third-party Help

  • Alice sends a box of items to Terrada Minikura sharing system
  • Terrada photographs and describes each item, gives Alice ledger entries (signed by them), and lists them Terrada website
  • Bob wants an item for a good cause. He trusts that one of Alice's items isn't junk because he trusts the information signed by Terrada
  • Bob asks Alice; Alice checks out Bob, says to Terrada, "Good, send it."
  • Terrada sends item, creates "gift" ledger entry for Alice and Bob
  • Bob receives/confirms item, creates "Thanks, it was great!" ledger entry for him and Alice
  • Bob has none of Alice's personal information, but Alice gets public credit

8 of 9

Kickstarter-type Example

  • Alice publishes entry, "Doing manufacturing run of groovy item; $10 to join."
  • Bob looks at Alice's history; sees she's done this successfully before
  • Bob pledges Alice $10, creates ledger entry, sends money now or later
  • On receipt, Alice creates ledger entry "I received $10 from Bob."
  • Alice does manufacturing run, sends groovy item to Bob
  • Bob creates ledger entry, "I got my item! It was great!"
  • This whole history is now public; Bob's friends now can trust Alice via Bob's evaluation

9 of 9

Key Points

  • This is not a particular program; this is a protocol
  • It stores statements made by people and confirms who made them
  • System is fully distributed; anybody can store/share ledger entries anywhere:
    • On local hard drive, on cloud drive, as plain file on website
    • On dedicated website that might provide other services
    • Complete set of entries for a particular evaluation can be built from multiple sources
  • Supports anonymous/pseudonymous operation, or "real" identity info
  • Like the real world, supports participants evaluating others on:
    • The particular criteria that are important to them personally
    • Using information from the ledger and other sources
    • Giving different amounts of trust to different information depending on source
  • Lots of different possible applications and multi-participant interactions