While the report One Hundred Year Study on Artificial Intelligence (AI100) provides a sufficiently comprehensive discussion on many aspects of AI's potential impact on the society, I would like to challenge the perspectives and discussion on how AI will impact education in the near future. The report provides a too positive view on how much AI will help improve our education quality and simply attributes the inefficient use of AI to the lack of financial resources.

The report heavily implies that offering education at scale is beneficial to the society, but in reality, it may not be true. According to the report, AI has enabled many instructors to enlarge their class size to hundreds or thousands (p.32). But the question is how effective is this type of education? Does this type of education really motivate the students to think and grow or only force them to memorize the correct answers? For individuals who are clear with their goals and purposes, it might be more efficient for them to learn using AI, since they are already motivated and know what they need to learn. However, for the others, learning by reading info from Wikipedia and doing some auto-generated multiple-choice questions might not be helpful. It is crucial for the instructors to offer mentorship on a personal level to guide the students. Education is not a commercial good, it is about helping individuals to explore their potentials while learning technical knowledge. Therefore, cautions should be taken when use AI to enlarge the class size.

In addition, the report emphasizes how AI can provide personalization to enhance our education at all levels (p.31). First, personalization is a very broad term and it can take different forms and depth, so I am not sure how personalized the learning process can be. Also, how do we evaluate the benefits and effects of the personalization provided by AI? How do we know whether the personalization is the best for the students to achieve their learning goals? Different individuals have different sets of learning habits, which makes personalization a complicated process. The report claims that AI can provide personalization at scale. Almost everything works at scale dont work for some special situations. The personalized learning plan would not be an issue for those who benefit from it, but issues may arise for those whose needs are not met by the AI personalization. Another issue with personalization is it may not be standardized, which could enlarge the gap, either in the specific knowledge or the amount of knowledge, among the students.

As the report mentions in the introduction, human brains are way much more versatile than the existing AI, and it is unlikely that AI will match human thinking anytime soon. Delivering education is fundamentally different from other tasks that AI is usually used for, as AI is teaching humans in this context. The application of AI in education may bring some benefits, but it may also backfire depends on how it is specifically used by the instructors and the students.  

