An AI system used for recruitment might favour applicants who speak English fluently or live in Kathmandu over equally qualified applicants from remote areas who speak Nepali or local languages. This affects the underprivillaged people in getting similar employment opportunities in Kathmandu. Therefore, human intelligent or business owners to be mindful regarding recruiting the employee based on english language competencies.
Your example made me think about how the recruitment algorithms prioritize English fluency as an inferior quality. I think another group of individuals who may be underrepresented by your algorithm are talented graduates from rural public schools who may not have had the same English training as their privileged counterparts. I agree with your point, but I am curious about how HR managers can be trained to interrogate the language models’ decisions before making hiring decisions.
I agree with your points. I would like to add that local teachers and communities should also participate in developing AI systems. They understand students' languages, cultures, and real classroom situations. Do you think including more local Nepali examples in AI training could make digital learning more inclusive and effective?