In recent years, Artificial Intelligence (AI) has become such a hot topic that anybody even with the minimum interest in technology industry seems should have heard about it at least once in their life.  It is true that AI is increasingly becoming prevalent in our society, and more substantial usage of it is expected in the future. According to the article “Stanford One Hundred Year of Study on Artificial Intelligence”, AI would not involve into dreadful threat to our life but rather serves for human good in longer term. I have no intent to negate such aspect, depicting horrifying Hollywood-like scenarios of AI completely taking over the role of human. Neither do I want to challenge the private security in AI technology, since in my own perspective such problem could be adequately solved, as solid policies are implemented accompanied with AI’s development itself. What this essay going to challenge is rather a nuance problem relatively ignored in the article: the inequality that AI would bring to overall human society. Since it is quiet broad topic, I would mainly focus on discussing about education domain.

One assumption underlying my argument is that almost every technological improvement has made our world be more and more polarized. It is to say, while the advent of fancy technologies undoubtedly provides millions of people convenience in their life, it also drags certain groups of people who are unable to get access of those technologies into a disadvantaged position. In other words, our current technologies are paradoxical in their existences that while they make more and more people feel closer, they also automatically get certain people segregated, and the two tails are getting more and more remoted as technologies getting advanced. Same situation happens on AI as well. In education domain, the study panel are expecting to see AI provides more personalized education to each student, since the emergence of massive scale of online courses guaranteed ample data sets to be collected. However, the question is to what extent are those online resources prevailed in various areas? And in turn how much would those personalized education courses fit to students if they are out of consideration even at the beginning when those AI algorithms are invented?

To be more specific, in the article, it says online resources would offer people opportunities for better education in the area where the broad population are facing difficulty to get education. Note the article also points out that this could happen only when the population have the tools to access those resources. But what is the possibility that people can get tools for online courses if a wild range of population are even hard to obtain fundamental education? Admittedly, there exist several areas like mid-west America where students find it more suitable for homeschooling, and online courses advanced by AI algorithms would benefit them without doubt, since AI enables teaching to be more personalized to its target. However, in most of the cases, people unable to get education are less likely to even get in touch with such online resources. Not to mention personalized education developed by AI, in some remote rural area of China, where internet is not even guaranteed, local governments are struggling to provide children with most fundamental education infrastructures. Students in such areas hardly know about iPad, which currently a lot of education apps are developed based on it. The major concern is, before those children even get access to the online education, they have already been excluded by training data sets for AI development. If this is the case, how can people claim AI enhanced education by providing more specialized schooling?
