This project is being conducted by a team of researchers to investigate barriers NLP researchers face in using and developing large language models (LLMs), as well as the conceptualization and theories of democratization in the context of LLMs. We seek to analyze where the goals of democratization researchers meet and diverge from the realities of NLP researchers in the Global South. We may follow up with individuals for paid in-depth interviews.
You can benefit from participation in this survey
- via the opportunity to investigate the opportunities/shortcomings of LLM democratization efforts, and
- by raising awareness in the community about the benefits/risks of current conceptualizations of democratization and the potential for more democratic technologies.
We do not foresee any physical risks associated with this study. There may be psychological risks given that some interview questions are linked to participants' experiences of marginalization.
Filling out this questionnaire is completely voluntary and there are no geographical restrictions to participation. You don't have to answer any question you don't want to, and may stop participating at any time without penalty. This questionnaire is non-anonymous as we ask for your email address to potentially follow up. Survey responses and e-mail addresses will only be accessed by the authors of the project, and will be deleted at the end of the research project, or upon request. Authors will NOT be compensated for their participation in this questionnaire, but will be compensated if they are selected to participate in a follow-up, in-depth interview. Selection to participate in a follow-up interview is NOT guaranteed.
This study (application number 23-10-8) has been approved by the Ethical Review Board of the Department of Computer Sciences at Saarland University.
The researchers conducting this study are:
Arjun Subramonian, UCLA, arjunsub@cs.ucla.edu (Point of Contact)
Vagrant Gautam, Saarland University, me@dippedrusk.com
Zeerak Talat, Mohamed bin Zayed University of AI, z@zeerak.org
