In Robert Gehl's chapter, "The Archive and the
Processor" he discusses the Amazon Mechanical Turk, which is marketed as
"artificial artificial intelligence". On this website businesses can
ask people to work on "human intelligent tasks" (HITs) and anyone can
get paid to complete these HITs. This website depends on microlabourers and
Gehl only mentions how it is used to process massive amounts of data. Since
2011, however, MTurk has had significantly more HITs based on being a subject
in psychological studies.
Issues
surrounding psychology's new dependency on the website include a lack of
knowledge about the actual workers, which can lead to inaccuracies and lack of
diversity in research. On one hand, it is possible for the identity of subjects
to be revealed, which is an ethical issue and a breach of privacy for the
worker. On the other hand, volunteers for MTurk can purposely distort
information they are providing, which would skew data for researchers. Microlabourers
may gain a sense of autonomy by being able to dictate their own working hours
and being rewarded for their work, but using a profit based website for the
advancement of psychological research can pose problems for both the subjects
and the researchers.
This is an article which further elaborates on the use of
MTurk for psychological studies:
http://nymag.com/scienceofus/2016/06/psych-researchers-might-have-an-amazon-mechanical-turk-problem-on-their-hands.html
This is an article outlining some of the issues with being a
worker on MTurk:
https://www.zerofox.com/blog/five-things-to-consider-before-using-mechanical-turk/
No comments:
Post a Comment