Human Powered Rails: Automated Crowdsourcing In Your RoR App
summary
Builds a human-powered API in RoR that automates background jobs to post tasks and import answers from Amazon Mechanical Turk, using the gems Turkee and RTurk. Demonstrates a live run: 21 Instagram sandwiches, 2 assignments each at $0.15, finished in 23 minutes with median 42s per hit and high consensus, plus a reprocess/adjudication flow for disagreements. Shows how inputs/outputs, validators and formats scale, and raises ethics and wage concerns including Cambridge Analytica’s MTurk use.
AI-written summary · about
chapters (52)
- 0:00Introduction
- 0:42MTurk
- 1:22Thanks Rails
- 1:52I owe you
- 2:09You probably wont leave
- 3:06Personal Growth
- 3:25Fake It
- 3:49Mechanical Turk
- 4:00Turk
- 4:31The Turk
- 4:47The 4 main use cases
- 6:01Making money as a Turk worker
- 6:59Turk tracking
- 7:17spammy
- 7:33fashion items
- 8:02use case
- 8:37french fries
- 9:00without Rails
- 9:53GUI
- 10:10Instructions
- 10:32Template
- 10:45Cost
- 11:12Output
- 11:48Time
- 12:15Results
- 13:08Automation
- 13:45Gems
- 15:03Batch Items
- 15:25Turkey
- 15:40Turkey schema
- 16:32Defining the model
- 16:56Form URL
- 17:10Turk Method
- 17:29Turk Form
- 18:07Import Result Data
- 18:44Batch Item Results
- 19:38Batch Item Processing
- 20:08Queue for Reprocessing
- 20:25Adjudicator
- 20:43Rake
- 21:17cron
- 21:32Use Cases
- 22:01Extending Crowdsourcing
- 25:13Output of Crowdsourcing
- 26:04Input Formats
- 26:46Twilio Bed
- 27:22Turk Tips
- 29:09Turk Ethics
- 30:02Turk Culture
- 30:41Cambridge Analytical
- 31:23Conclusion
- 32:06Outro
