~
Urbit

Communal Computing on Urbit

User research, domain research/synthesis, white paper creation and publicity effort around Urbit ID as a solution to questions of IoT identity sharing and privacy

February 17, 2022

Reward: 4.6 stars

ID: P0113

Grantee(s): ~pilwyc-fastec

CompletedProposalOther

Communal Identities for the Urbit Smart Home

Today’s smart home, IoT, and other communal devices should be for the people living in a home but are really surveillance systems for large corporations and governments. In this project we will make Urbit the best OS to tackle the identity, security, privacy, experience, and ownership issues with communal computing.

Overview

There are a record number of devices being placed in the home and office that are meant for everyone to use. This includes smart home devices like video doorbells, internet of things (IoT) devices like light bulbs, and home assistants like the Amazon Echo. However, they are framed and built wrong due to their focus on a single user context from the days of mainframe time-sharing. They are built for the person that sets up the device rather than all of the people that may have access like partners, children, and guests. This causes problems with identity, privacy, security, experience, and ownership.

For example, the Amazon Ring video doorbell or connected security camera collect lots of video information. It attaches it to the identity of the owner and then makes it available to local police forces. It doesn’t ask anyone that comes into view to opt-in or consent to having their image captured. There are many privacy and safety issues with automatically collecting this information without consent. There are reasons that someone would want to be identified though like a member of the home that would automatically be let into the front door when they are seen. This isn’t the same as the random delivery person leaving a package as their identity isn’t needed for the task of detecting a package is there.

In Urbit, the identity model (Azimuth) is key to running the OS and it has a great opportunity to rethink the paradigms of communal identities. The strong identity model for Urbit is beneficial but needs to figure out how to allow for other people besides the owner of the device to use the home’s infrastructure.

From an identity point-of-view, we are made up of many different pseudo-identities: as part of a family, a work organization, a set of friends, a hobbyist community, and a neighborhood. All pseudo-identities that make up a person are interlocking with other people’s identities through association (aka p2p). This includes a family member, a co-worker, a friend from school, a guest in your house or your nextdoor neighbor. This leads to the intersection of various reputation systems that overlap with Urbit and other reputational institutions (some of which may not exist yet).

I propose that I embark on an ethnographic study of the intersection between Urbit, identity, IoT, and smarthomes. Then create a whitepaper describing a way forward for Urbit.

Scope

We will complete the proposed project with the following artifacts:

  • Ethnographic research synthesis - and potential prototypes used to discuss

  • Whitepaper describing how communal devices could leverage Urbit - this will include an outline of key problems identified during research, Urbit use of identity for communal devices, any Urbit feature recommendations, and a concrete case study like an open source video doorbell

  • Blog post summarizing the whitepaper

  • Talk and workshop regarding the whitepaper open to the public

To get to these artifacts the following work will be done:

  • At least 10 compensated interviews with Urbit users or similar proxy profile focused on IoT and smart home - see the appendix interview guide for more information

  • Prototyping sessions to show/discuss with interviewees to make their feedback grounded in a concrete case - this could point to an additional round of interviews after the initial ones

  • Syntheses sessions to pull themes out of the research

  • Ideation sessions to consider various IoT and smart home scenarios

  • Technical sessions to discuss identity and Azimuth

  • Design sessions for concrete case studies

  • Drafting, review, and editing of whitepaper, blog post, and talk

All work is open to developers and other members of Urbit.

Key scenarios for the identity model with regards to communal use cases which may expand during research:

  • Adding comets for devices to push content to the planet (or other storage locations and APIs)

  • Auto-provisioning (and collecting consent from) identities for joining the home (both people living in the home and guests) to comets

  • Auto-de-provisioning identities when people leave the home (and removing other data collected for their identity)

  • Attaching biometrics to identities for automation

  • External reputation system integration for local governance

  • Existence of Urbit-based IoT alongside other smart home devices

  • Linking current smart home platforms (e.g. Google Assistant and Smart Home) to Urbit IoT devices - for example, to allow a video doorbell to show it’s video feed on a Google Home Hub

  • Use of DID (and VCs), SSI, W3C (e.g. WoT), and other identity, IoT, etc. standards to integrate with next generation of identity services

My experience and why I’m the right person for this project

I have over 20 years of product management experience in building B2C and B2B products, apps, services, etc. This includes product and business development leadership at Microsoft, Waze, KAYAK, Facebook Reality Labs, and most recently Cognizant. I’ve also founded a startup in the restaurant tech space, ran innovation for a venture capital firm, and directed AI/ML research projects for a design consultancy. I have experience in starting projects from zero and taking them through research, deployment, and maintenance. This happens because of my focus on connecting execution to a strong strategy and future vision.

Regarding communal computing, I’ve written and evangelised the concept of as a way to solve the current problems with home and office devices (see the appendix below). I’m an IoT hobbyist, smart home enthusiast, Alexa skill builder, and most recently was product manager for the identity platform, biometric recognition machine learning, and voice assistant integration for Facebook’s Portal.

I’m interested in finding models that make sense for the smart home that don’t give all of a person's data away and give them the control to keep it working the way they want, regardless of company boom and bust (or M&A) cycles. Google Docs version with links↗

Milestones

Milestone 1: Research

Payout: 1 star and .15 WSTR Due Date: 3/11/22

The first milestone is after research into the problem area of communal use of Urbit systems. This includes interview guide drafting, recruiting, interviewing (at least 10 45-minute interviews), analysis, and synthesis of research.

The handoff is a research report (3-5 pages) with themes and important nuggets (~50 nuggets).

As compensation for the interviews we will give out specific Goras to people that participate in this study.

Milestone 2: Concept Development

Payout: 1 star and .15 WSTR Due Date: 4/11/22

The second milestone is to turn the research into actionable recommendations. This includes working sessions (at least 5) with the Urbit community to refine problems, ideate, design solutions, and situate in the context of Urbit’s current systems. This will include 5-8 people in a synchronous meeting to workshop.

The handoff is a compilation of ideas with any recommendations from the working sessions. This will most likely be a journey map format with problems and possible solutions.

Milestone 3: Report Creation + Sharing

Payout: 2 stars and .3 WSTR Due Date: 5/27/22

The final milestone is the completion of a whitepaper (10-20 pages). This includes the whitepaper drafting and editing. As well as, authoring a blog post (1,000-2,000 words) and live talk for the Urbit YouTube account (45-60 minutes).

This will be shared out with the wider boomer web distribution like blogs, podcasts, newsletters, conferences, meetups, Twitter, and LinkedIn focused on topics including identity, privacy, crypto, IoT, and smart home (10-20 to be identified).

Here is a short list of possible groups:

  • O’Reilly Radar blog
  • Building the Metaverse Medium publication
  • Inside Blockchain newsletter
  • NYC Media Lab Data Download newsletter
  • The IoT podcast
  • VerseTech SF Metaverse | AI | Blockchain | XR VR AR meetup
  • Vienna Digital Identity meetup
  • Privacy Engineering SF meetup
  • MLUX meetup
  • IIW unconference