WinterMilestone 1 Duka
Da Eun Sally Chung Lee: @sallyxchung
Nicolas Rodriguez: @nick13rodriguez
Samarth Sandeep : @samarthsandeep
Rhonda Armstrong: @rarmstrong
Experience the Life of a Worker on Mechanical Turk
-Simple lists, which makes it very direct.
-No hierarchy of information to show what is important.
-Each HIT provides instructions, explaining the task.
-The worker is able to track their progress.
-Easy tasks can be earned, keeping an incentive.
-Tasks can be enjoyable and interesting thus share various fields (can learn from tasks).
-The platform seems that it could be improved. The design was something I did not enjoy in particular. The design of it could be more user-friendly – it unmotivated to enter the site.
-Difficult to communicate with the requester.
The platform is not available globally, making the online workforce limited.
Sign up process can definitely be difficult if you need to create an Amazon + Amazon Payments account (may require ID fax sent for further verification)
You may not truly be qualified to complete a HIT (Requesters place further barriers)
need some way to better the search and filtering system
Task details can be very vague
There are spammer requesters
Terrible UX; Accepting a task is not as simple as clicking on a task and hitting accept
Very bland; Has all of 4 colors, and uses quite dull typography
it doesn’t feel like a workplace: there is no coworking, no sense of individuality
minimum wage/labor laws cannot support workers thus workers be taken advantage of
Experience the life of a Requester on Mechanical Turk
Strengths: Direct set of categories so that requester can visualize the data. Tasks are very straightforward. Gives details of the task Makes finding a random sample group simple Payment onboarding is simple and does not require many steps
Weaknesses: Design UI inconsistencies Some confusion with using it the first time – had to play around it to get used to it. There can be barriers how of requesters describe the task, meaning workers may not understand fully Collecting data is a little too simple; could use deeper analytics for deeper connections QA is nonexistent; you have to check each piece of work you get, and approve it requesters should be able to select a targeted audience among workers (ie. based off simple demographics) there needs to be some sort of verification process to combat spammer requesters. should be easier to send follow up tasks to workers that have done some of your HITs mturk needs some a message system
Explore the crowd-labor markets
Strengths: Brings knowledge from different perspectives into one. Perfect knowledge. Creates globalisation Creates easy, accessible information online Very modern Efficient, accurate Closed community, but not really Tasks are enjoyable Can be a way to accumulate ‘rainy day’ funds (amazon account - mTurk) Opens tasks to a group of people that are new to you, rather than to family or friends Provides new sources of income for people Very decentralized; kinda like the blockchain of work, so it can be done anywhere
Weaknesses: How are we going to be able to know whether the person that uploads information is correct/valid? – Trust issue Dependency towards technology — leads to less physical labour - reducing jobs - creating economy problem More jobs online - creates health problems - eye strains, less active, The issue with sustainability: what are we going to do with old phones? Society becomes too specialised Clashes between members Doesn’t teach anything; builds on top of existing knowledge, rather than allowing one to start a task and have the ability to learn a new skill from that Makes people spend hours in front of a screen; not very healthy for day-to-day life very antisocial; removes the need for one to learn how to write good emails or participate in small talk has no worker benefits besides pay; no insurance, stock options, etc has no unions or governing bodies for workers
MobileWorks is a crowdsourcing platform that is based through mobiles and it aims to allow the less advantaged people in society to be able to get involved in society through the crowd source industry.
Helps low income earners Simple and effective Could help all classes in the economy Promotes equality by giving everyone an opportunity Uses OCR; helps to spread new ways to do old tasks
People become dependent on their phones leads to greater health risks Difficult to trust The work could end up being unprofessional since people are working through phones Tasks could be limiting - no audio tasks UI is quite primitive; makes the worker feel as if they are working alone and only for short exchanges, rather than for a full day Payment gateway nonexistent; would still rely on older systems, such as sending cash by mail or by person, and thus, inhibiting the promotion of banking systems and mobile wallets for future pay
simpler UX: does not need so much white space and lettering use OCR to do OCR: allow the user to write down the solution to the task and take a picture of it better payment gateway: use blockchain or mobile wallets to allow for easy payment and to promote decentralized banking
Daemo is an employing open community crowd sourcing platform.
Strengths: Easy to use Development of trust through feedback such as ratings Community feel More focused with the requester and worker Feedback loops allow for greater interaction before a rating is made rating is also updated based on whether the project is on schedule or not restores trust: builds on human connections, rather than task connections
Weaknesses: Clashes between workers - since it is a group collaboration Still doesn’t address societal shifts: while it does make crowdworking better and more appealing, it still does not address income, education, and other key problems that crowdworking could play a key part in solving
Improvements: Machine learning to sort tasks: use adaptive resonance to link tasks back to the worker’s initial qualifications to find the best fit easily chat rooms: allow for workers to chat with requesters about possible tasks, rather than solely having set tasks, to promote more of a workplace environment suggested lessons: use language processing to find an edx course that matches the skill required for a task to allow for someone to educate themselves either to complete that task or for their own knowledge
Flash Teams are guided teams of crowd experts.
Strengths: Easy to manage - making tasks to be completed faster rapidly coordinates large groups of people Puts task management on a system that reflects time uses lo-fi wireframes to help get people without design expertise participating in workflow
Weaknesses: focuses more on the requester than the worker: does not incentivize workers to continue using the platform does not solve language barriers: still makes users all type and read in English Is very narrow; focuses almost entirely on design More like group freelancing than a platform that actively helps to change the face of work
Improvements: broader tasks: have tasks beyond app dev and design use moxtra’s draw sdk to allow workers on a team to draw and communicate prototypes with each other employ a central translation application: have tasks appear in a different language for people working in another country, and work documented in the same language as the requester to allow greater globalization; use more icons to represent tasks