URGENT: Daemo at a European Dialogue on the Platform Economy Event, Brussels

[neilthemathguy]

My Proposal

For the a European Dialogue on the Platform Economy at Brussels, the organizers have asked for Daemo's abstract. At present, they have added the abstract from the Daemo's vision paper Gaikwad et, al. 2015.

I've updated the abstract to cover various aspects of Daemo. Feel free to add more. We will need to send this to the organizer as soon as we can. Due to the urgency, I am marking this as an URGENT proposal, deadline Sunday 1/21 11:45am EST.

Daemo

Paid crowdsourcing platforms provide opportunities for autonomous and collaborative professional work as well as social engagement. However, in recent years, the lack of trust and uneven distribution of power among platform teams, workers, and requesters have threatened the existence of these platforms. To address the challenges of trust and power, the Stanford Crowd Research Collective, a group of worldwide researchers has developed Daemo, a self-governed crowdsourcing platform. Daemo, an open sourced research project, is distinguished by its three fundamental building blocks: Boomerang Reputation System and Crowd Guilds, Prototype Tasks Design Workflow, and Governance Process.

(1) Boomerang Reputation System and Crowd Guilds: while crowdsourcing platforms suffer from reputation inflation, Boomerang aims to elicit more accurate feedback by rebounding the consequences of feedback directly back onto the person who gave it. With Boomerang, requesters find that their highly-rated workers gain earliest access to their future tasks, and workers find tasks from their highly-rated requesters at the top of their task feed. Inspired by a game-theoretic notion of incentive-compatibility, Boomerang opens opportunities for interaction design to incentivize honest reporting over strategic dishonesty. Daemo also proposes Crowd Guilds, centralized groups of crowd workers who collectively certify each other's quality through double-blind peer assessment.

(2) Prototype Tasks Design Workflow:
Workers and requesters are often unable to trust each other’s quality, and their mental models of tasks are misaligned due to ambiguous instructions or confusing edge cases. To date, workers have been blamed for low-quality results; should requesters also bear responsibility for low-quality results? Daemo's Prototype tasks design workflow encourage requesters to take ownership of the negative effects they may have on their own results. In order to fix downstream impacts of requesters' decisions on work quality, Prototype tasks asks a small number of workers to provide a feedback so that requesters can revise their instructions, payment, and task designs before launch.

(3) Governance Process: Daemo's constitution aims to provide platform team (Stanford Crowd Research Collective), crowd workers, and requesters with equal ownership and a means of governing the development and evolution of the platform into the future. The constitution outlines the goals of the platform, relationship between members of the community, and methods for seeking ideas, amending the constitution and resolving conflicts. With these three fundamental building blocks Daemo's research aims to reestablish the trust in crowdsourcing platforms.

Daemo
https://www.daemo.org/home
https://www.sharersandworkers.net/wp-content/uploads/2018/01/daemo-uist.pdf

Prototype Tasks Design Workflow
http://bit.ly/2n01QTV

Boomerang Reputation System
http://bit.ly/2Bh4zwo

Daemo Crowd Guilds
http://bit.ly/2DqO2Mu

Governance Process Daemo Constitution:
http://bit.ly/2EZDDDD

Implications

Short-term: Daemo's abstract for A European Dialogue on the Platform Economy
Sharers and Workers

Person or People Who Added the Proposal to GitHub

@neilthemathguy (Neil Gaikwad)

---

Use comments to share your response or use emoji 👍 to show your support. To officially join in, add yourself as an assignee to the proposal. To break consensus, comment using this template. To find out more about this process, read the how-to.

[qwertyone]

I was reading the proposed themes, and the prototype task authoring work flow angle will be needing some adjustments to fit their content and focus. The reputation system, crowd guilds and governance topics seem to be more aligned with their content priority in comparison. You need to keep in mind that Unionized Labor is big in the EU and currently there are strikes throughout countries as platform freelance workers (Deliveroo: http://www.telegraph.co.uk/news/2018/01/20/deliveroo-riders-go-strike-belgium-netherlands/ is the most recent) try to become recognized as contract employees rather than independent workers. To avoid these common European dynamics for example, Uber is attempting to appease this dynamic by offering free benefits (https://techcrunch.com/2017/07/25/uber-is-going-to-provide-accident-insurance-in-france-for-free/) to offset their drivers normal expenses; benefits that internalize costs to the employer is one of the major incentives for the free lancers to put pressure on these companies. This said, if you choose to keep the task prototyping in the first position of the talk, I think that it might be wise to might reshape the discussion to the interface and programming as a the first step to mediating the role of governance, as such the platform is a benevolent dictator controlling the options between parties. The reputation system Boomerang also falls within this thematic umbrella. Minimizing the conflicts between the outcome expectations of the contracting parties is a first hidden step to a comprehensive governance system. After all, people will rather work for companies they like rather than ones that unintentionally create conflicts through the platform.

[neilthemathguy]

Thank you Aaron (@qwertyone). Do you suggest any actionable changes to the Prototype Task description?

Above description is for the webpage. I can reshuffle the order. I'm planning to scope the talk around some of the issues you and organizers have highlighted above. I'll share the slides in Slack tonight.

[qwertyone]

I was thinking about the structure of information.

Programmatic

• Prototype Tasks Design Workflow
  -- (Caveat: If this angle is true from the studies.) It has been demonstrated that requester task authoring affects the quality of the work they receive. Unintentionally, they undermine their own work requests and blame the worker. We are working proactively to mitigate unnecessary dissatisfaction through the workflow so that workers can avoid unnecessary conflict and requesters hidden and preventable negative word of mouth on Turkopticon and REDDIT forums. We are still working to demonstrate this following statement, but we believe that by preempting conflict, we can retain a satisfied community, whom would be more willing to participate within Daemo's governance process. If people don't want to participate, they cannot be governed. (A king without a kingdom)
• Boomerang

Social

• Crowd Guilds
• Governance

[neilthemathguy]

I see, the talk can be organize to convey various perspectives. The description should be coherent to represent the platform. All interventions in Daemo have both Social and Programmatic aspects.

The Prototype Task design workflow is to create the shared mental model between workers and requesters. While it may help in conflict reduction, in the description of the platform we should not make any causal claims that are not proven/researched yet.

I have ordered the sequence and updated the description to reflect Workers and requesters are often unable to trust each other’s quality, and their mental models of tasks are misaligned due to ambiguous instructions or confusing edge cases.

[qwertyone]

I am glad it helped.

…

On 21 January 2018 at 14:54, Neil Gaikwad ***@***.***> wrote: I see, the talk can be organize to convey various perspectives. The description should be coherent to represent the platform. All interventions in Daemo have both Social and Programmatic aspects. The Prototype Task design workflow is to create the shared mental model between workers and requesters. While it may help in conflict reduction, in the description of the platform we should not make any causal claims that are not proven/researched yet. I have ordered the sequence and updated the description to reflect Workers and requesters are often unable to trust each other’s quality, and their mental models of tasks are misaligned due to ambiguous instructions or confusing edge cases. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#54 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFy3xvRIRhPL_2yUVPvkzitr86z3pjBXks5tM0GOgaJpZM4Rlf0X> .

[neilthemathguy]

Sent!

[qwertyone]

Thank you for the confirmation.

…

On 21 January 2018 at 18:02, Neil Gaikwad ***@***.***> wrote: Sent! — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#54 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFy3xpBq6khlVMGCYPPO_Hz8UsYvG7Smks5tM22egaJpZM4Rlf0X> .