Although the writer has the best intention to ensure the benefits of AI in education, we cannot confirm whether the predictions of the education system in 2030 are plausible or well-functioning as the writer wished. According to this report with a time span of 15 years, we have no direct evidence to make a conclusion that the class size will be doubled with robotics’ assistance, nor the class material would be more personalized from K-12 to colleges. Without seeing the whole picture, any researches or regulations made depending on this analogy would stir up inferior impacts for our future generation. 

The arguer’s reasoning linking the accessibility of online learning with the augment of class size seems reasonable on the surface, but it may not be true after further scrutiny. The possibility of automatically generating quiz questions or offering online recourse surely reduces professors’ liability. Questions randomly generated by machines, however, usually follow a pattern, which leaves a loophole for students to take advantage of. Students who score high in those questions might not fully comprehend the material. Although AI can teach people how to perform an act, it can not teach them the logic and reason behind it. How can a machine train human's mind when itself has no conscious? If one student is confused, one may experience difficulty seeking for personal help due to the reduced human resource. There is merit to increase the size of class for courses that require less logical thinking and mainly rely on memorization. Yet, people who want to pursue profound knowledge in higher education systems are unlikely to perform well in such setting. Without eliminating this plausibility, the author can not conclude that the augment of class size with AI’s assistance would be necessary or beneficial. 

Aside from the legitimacy of the course materials generated by AI, deploying AI in nationwide scale requires the curriculum to be standardized completely. If every person is presented with the same information, it would require all instructor teaches the identical knowledge in the identical ways. What if some teachers are good with using textbooks and others prefer assigning projects? It hinders educators from offering their special understanding of the material to educatees. Studying the same material from the same platform might deteriorate the variety of children’s learning methods. Having such a uniformed way of perceiving new information might create bias or blind spot for scholars to conduct independent research studies. In stark term, the government will be able to easily manipulate citizens’ perception of truth if everyone learns from an AI system. They can simply modify the material from a far distance. Therefore, whether AI can make learning more personalized must be based on a more thorough investigation. 

Finally, the writer supports the prediction by over-generalizing the responsibility of education. Not only students need to learn a set of things, they also need to know how to learn. For college students, acquiring the ability of critical thinking is far more important than studying a file cabinet for facts. However, the machine can not provide its thought process to learners. Without knowing how to learn, students would not be able to create new innovation. 

Interfering with the education system might cast treacherous damage in the long-term. Insofar, we cannot prove whether AI can increase the class size, make a customized profile or educate people the ability to reason. Little reasons we have to believe replying AI to teach students would have a positive impact on the community. Rather, the best solution is to conduct a full investigation with more consideration until we can see the ins and outs. 