WinterMilestone 1 vinayak

From crowdresearch
Jump to: navigation, search

This is my submission for Winter Milestone 1. Completing the milestone was a wonderful experience, it helped me gain insights into the crowdsourcing ecosystem that no amount of reading would have.

Experience the life of a Worker on Mechanical Turk

My registration for Amazon Mechanical Turk was not approved after 48 hours so I used Microworkers to experience the life of a worker. There were several issues that the workers have to face in the current crowdsourcing platforms.

The biggest hurdle in my opinion is the uncertainty of the submission being accepted and the workers being paid at all. @JSilver and I had an interesting discussion on this issue on the general forum of slack and that discussion is being condensed into a wiki for future reference. Some jobs have a waiting period of as long as 7 - 10 days, this really takes a toll on the workers. A more stable environment where they know the chances of their work being accepted and having decisions made quickly will help many workers plan their finances in a much more efficient way at the same time saving them the trauma of waiting indefinitely.

Another major issue is pricing. Some jobs have a price of as low as $1 per hour, in my opinion this kind of a compensation borders on exploitation of the workers. An elegant solution to this problem can be to fix minimum wages for each category of work for each country. These minimum wages can be voted upon by the community, drawing inspiration from official data from reputed sources like the world bank. Of course requesters will be allowed to pay more than the minimum wage. This will ensure that people get paid a fair compensation at the same time ensuring that requesters can get huge amount of work done for a reasonable price thus preserving the essence of crowd-sourcing.

Acceptance of tasks is also of concern. Requesters have been known to disqualify the work of many workers just to avoid paying them. This is completely unfair. A way to counter this could be to have another set of workers rate the original submission. If the requester now disqualifies a lot of submissions that have been highly rated by independent workers then he can be penalized in some manner (rating downgrade etc).

Many requesters have unrealistic requirements. For example below is the screenshot of one of the job requirement I found on Microworkers

Screenshot Requirement.png

Experience the life of a Requester on Mechanical Turk

The requester surprisingly also faces many issues on a crowd-sourcing platform

Quality of work submitted many a times does not meet expectations. This is especially true for tasks that have very specific requirements. This can be very frustrating for a requester.

Another major issue is communication. It is sometimes difficult to make the workers understand some complicated tasks that have to be executed in precise way. This again feeds into the problem of poor quality of work received by the requester.

Readings

MobileWorks

The authors identify an interesting segment which could provide a huge boost to the number of workers available at the same time providing a way for people to develop a secondary source of income especially to those who need it the most quintessentially capturing the win win spirit of crowd sourcing.

Though interesting, the application is increasingly becoming redundant at least in India as users shift to smartphones. In fact many of the new users are bypassing the feature phones entirely, just as they bypassed fixed land line phones. When a smartphone is available for as low as $25.13, the first phone that many people buy is smart.

Daemo too should be mobile friendly as statistics suggest that people are increasingly accessing the Internet using their smartphones in fact last year the the number of Google searches via a smartphone bypassed the number of queries received through the desktop. In fact I believe that Daemo being the future of crowd sourcing should openly embrace the ubiquity and diversity of mobile devices. We should be awesome on the smartphone but not stop there we should go further and make the platform available on Smart T.Vs, Virtual Reality devices, Google Glasses and Holo Lenses of the future. So that we equip the users to demand and cater to the tasks of the future which are not limited to translations and transcribings.

Daemo

The single biggest issue that current platforms face is that the workers do not trust the requesters and the requesters do not trust the workers, the Bomerang feature that Daemo proposes is just what is needed to help overcome this trust issue and help build mutual respect between the workers and the requesters.

An amazing user interface is also an important feature and will help more users adopt the platform.

The Prototype functionality is just what the doctor ordered. It gives the users a unique opportunity to calibrate themselves this will ensure that both the workers themselves as well as the requesters are not disappointed in the future.

After the initial launch we can explore the possibilities of opening the platform to languages other than English. This will help overcome the language barrier and help more users to adopt the crowd-sourcing platform world wide.

Flash Teams

This platform allows people to use their talents and does not rely on basic human capabilities. Workers can earn according to their skill set, this not only helps the workers but also allows the requesters to benefit from specific tasks that cannot be performed by everyone at the same time promoting skill development. The modularity of the system and ease of communication also argue well for the future as already noted these are some of the major problems in the current systems. Pipe lining will help increase collaboration. Choosing and filtering experts can present a challenge to this model

Milestone Contributors

Slack usernames of all who helped create this wiki page submission: @vinayak