A Mechanical Turk Worker\'s Perspective

May 22, 2017 | Autor: Kristy Milland | Categoria: Sociology of Work, Work and Labour, Crowdsourcing, Amazon Mechanical Turk, Digital Labour, Crowd Work
Share Embed

Descrição do Produto



A Mechanical Turk Worker’s Perspective I have been a Mechanical Turk (MTurk) worker, known as a Turker, and community manager of the oldest worker forum, TurkerNation.com, for more than a decade. The site we workers call MTurk has changed over the years from an experiment in what work could be completed through crowd work to a race to the bottom experiment in how little one could pay for said work. Today, MTurk is a place where there are few rules: You can reject any work you wish to, pay as little as you desire, and never even talk to the people who are doing your work for you. It is clear that Amazon did not build promoting ethical worker treatment into MTurk; however, through social pressure on those who use the system to get work done, or creating new platforms that promise workers a better deal, we can turn the tide of crowd work to be in favor of the worker. When discussing ethics on a work platform, pay is always going to be top of mind, but ensuring workers are paid fairly is next to impossible on MTurk. To begin, what is ethical pay? Is it minimum wage in the United States, where Amazon is based? Or is it minimum wage where the individual lives, which is significantly higher in the province I live in than the U.S. federal mandate? What about workers in countries with no minimum wage? Even if we pick a country and go with that rate, how do we calculate minimum wage when people are doing piece work from home? As an example, another Turker and I tested how many identical pieces of work, known as HITs, we could do within an hour. I was able to complete 600 while she could only do 20. There are myriad reasons why we had this disparity, but we were both native English speakers with postsecondary education and adequate computer and Internet connections. In this case, should I be paid only minimum wage for completing 30 times the work she did? If even calculating a fair hourly rate is this difficult, it becomes an impossible task to ask Amazon to enforce a rule on minimum pay. Another consideration for pay might be to ask for specific per task minimum rates, but often this leads us down a path to racist conclusions. For example, just because a worker who does not speak English natively can easily transcribe a receipt, some have said we should only pay what is considered a fair rate in a country such as India (Infoscout, 2016). Does that mean a worker in the United States doing the same work should only be paid a few pennies? And what of Indian workers who happen to have advanced skills that allow them to complete a complex task? Are they automatically only paid a few dollars an hour because that is what the requester of the work feels is fair in that country? Even setting a minimum per task rate becomes extremely complicated when you are trying to determine how half a million people from around the world with a large variety of skills, educational achievements, and desires deserve to be paid. Forcing Amazon to set some arbitrary minimum rate may benefit workers, although I question whether it would increase pay or instead just increase the size and scope of each HIT posted. I think we would be better served to focus on one of two alternate options. The first is pressuring requesters of work to consider ethical pay and treatment of workers before they post jobs to be completed. Amazon could have a part in this, updating their guides to include information about how to contact the various Turker communities to get input on factors such as pay rate, HIT design, and getting better results. By redirecting the requester to the workers, it is far more likely the advice they get will be in the workers’ favor and, therefore, likely more ethically biased than the advice they get from other requesters or Amazon themselves. This would also push requesters to communicate with workers, something that only happens rarely today as Amazon puts no pressure on them to answer e-mails or reach out to us at all. This would confront the requester with the fact that workers are human beings, something so easily forgotten when we are represented by an image of cogs in a machine and referred to as Artificial Artificial Intelligence by Amazon itself (Amazon Mechanical Turk, n.d.). Humanity breeds empathy, and with that requesters may no longer see us as software at their bidding, but instead as laborers who deserve as much to be paid fairly as the requester might feel they themselves do.



The second option, which I currently advocate, is to create worker-run platforms, either along the lines of a cooperative or some other format where workers have a say in the structure of the labor market. An example of this is the Daemo marketplace, which is currently in beta testing (Gaikwad et al., 2015). Daemo is a crowdsourced crowd work platform with hundreds of participants helping to make it a reality, including many workers. Other projects may be launched after lengthy discussions at the recent Platform Cooperativism conference (Platform Cooperativism, n.d.), where worker input was valued and cooperative platforms with workers at the helm were dreamed of. Regardless of who may participate and in what format the platforms are to be created, the most important action we can take is to move forward. Let us give workers new venues where ethical treatment of all parties—workers, requesters, and those who may staff the platform itself—is at the forefront. Although it may be tough to sway the actions of a juggernaut such as Amazon, we can alternatively fight back by creating new venues that offer what their platform is lacking. This may be the incentive needed to push them to reconsider ignoring the workers’ plight. References Amazon Mechanical Turk. (n.d.) Retrieved from https://www.MTurk.com/MTurk/welcome Gaikwad, S. N., Morina, D., Nistala, R., Agarwal, M., Cossette, A., Bhanu, R., … Mithal, A. (2015, November). Daemo: A self-governed crowdsourcing marketplace. Proceedings of the 28th Annual ACM Symposium on User Interface Software & Technology (pp. 101–102). New York, NY. Infoscout. (2016, May 11). @Rochelle In India that’s more than what most people make in a day. We’re providing them a better option than what’s available locally. Retrieved from https://twitter.com/InfoScout/status/ 730452372514689024. Platform Cooperativism: Nov 13–14, NYC. (n.d.) Retrieved from http://platformcoop.net/.

Kristy Milland TurkerNation.com [email protected] © 2016 Taylor & Francis http://dx.doi.org/10.1080/23736992.2016.1228813

Pragmatism and Mechanical Turk: Citizenship and Labor Rights in Digital Communities of Knowledge Researchers should see the digital worker as a citizen, imbued with all the rights and protections the Fourteenth Amendment calls for. In the spirit of pragmatic social science, this commentary proposes that the Fourteenth Amendment catch up with the proliferation of digital labor. A Deweyan approach to citizenship can be helpful in solving the issues of those digital workers that are exploited. In other words, if the digital worker is granted protections as a member of a global digital community, then, in a Deweyan sense, their crowdsourced labor can be seen as community service rather than a series of simple economic transactions. Almost a century ago, the Fourteenth Amendment was being interpreted to extend citizenship rights universally, to even include corporations. Philosopher John Dewey lauded dissenting Justice Oliver Holmes in 1926 for arguing that the amendment should be interpreted with an emphasis on social experimentation and problem solving as opposed to dogmatic and reductionist reasoning (Dewey, Boydston, & Sidorsky, 2008). The world is a complex and beautiful place abounding with rich potential, says Deweyan Pragmatism; however, there is a tendency to use information to mechanize the lived environment, to reduce it to a simple cause-and-effect model. The Pragmatic solution is to employ scientific inquiry to enrich the lived experience, to favor the creative over the habitual (Diggins, 1995).

Lihat lebih banyak...


Copyright © 2017 DADOSPDF Inc.