|
|
| Line 4: |
Line 4: |
|
| |
|
| == Technologies == | | == Technologies == |
| '''1. MTurk-centric tech innovations'''
| |
| :Import badge concept from Amazon retail store; add filter option to highlight badged HITs
| |
| ::Real Name
| |
| :::Opt-in , linked to credit card of requester
| |
| :::All HITs from that requester have Real Name badge
| |
| :::Goals
| |
| ::::Badge increases accountability through attribution
| |
| ::::Filter increases difficulty for unbadged HITs to attract workers, incentivizing requesters to opt in.
| |
| ::"Real Purpose" + filter (still foggy)
| |
| :::Opt-in
| |
| :::Separate "purpose" dialog box on HIT preview page with "Does this seem credible?" button
| |
| :::HITs with a certain percentage of "yes" responses get "Real Purpose" badge
| |
| :::Goal: Badge increases disclosure and decreases alienation from work; Filter encourages growth of disclosure norm.
| |
| :Import Review system for HITs from Amazon retail; add "highest rated" to sort-by drop-down
| |
| ::Character limit to reviews
| |
| ::Include "Was this helpful" or Urban Dictionary-esque "thumbs-up, thumbs-down" to evaluate the reviews
| |
| ::Supplements Turkopticon's requester-based review plug-in, which can be circumvented by requesters having multiple accounts.
| |
| ::Goal: Add interactivity and immediacy to current feedback mechanism
| |
| ::Question: Does this overlap too much with/obviate need for Real Purpose?
| |
| :Add social features for sense of community
| |
| ::MTurk Livechat
| |
| ::Simple box in corner of MTurk interface
| |
| ::See Gwap for example
| |
| ::Requester profiles: Picture, location, business aims?
| |
| :Minimum wage?
| |
| :Challenges
| |
| ::Can all these features be implemented as external plug-ins? If not, how could Amazon be convinced to add features to a purposefully light, streamlined, "non-human" platform?
| |
| :::Social capabilities (counterintuitively) increase aggregate productivity by encouraging Turkers to 'hang out' on the site
| |
| :::Review system gives MTurk more long-term legitimacy as a site that cares about its workers
| |
| ::For the features that can be implemented as plug-ins, how to get a critical mass of people to install the plug-in? I.e. how to avoid Turkopticon's apparent sparkle-and-fade?
| |
| '''2. More general tech ideas'''
| |
| :Non-profit/philanthropic platform (Extraordinaries)
| |
| ::Alternatively, add "volunteer/philanthropic" to filter option in MTurk. But how to get Amazon to do this when they have zero monetary incentive (no cut for unpaid tasks, right?)
| |
| :Reputation Portability
| |
| ::OpenID for digital labor platforms, where user identities come with basic reputation information?
| |
| ::Something involving microformats, to make reputations "machine-readable"?
| |
| ::Challenge: how to convince digital labor platforms that interoperable log-ins, etc. won't hurt their market share
| |
| ::To do: E-mail Chris Messina
| |
|
| |
|
| == The Worker-Owned Platform == | | == The Worker-Owned Platform == |