Learning using AI and computers in general tends to be cutting corners and diminishing the effort one has to do. Because of it the learner has less direct experience with the thing he learns, which means he doesn’t learn as well. Maybe he notices he understands the subject, but the question is: will he remember enough of it? Will he actually “gain” something with it?

I have been learning Japanese by using software, offline and online, and although the learning process went smooth, my feeling was that I hadn’t learned anything which helped me in real situations.

Obviously, real situations are less standardized than a computer setting tends to be, and that’s why there isn’t enough variety in the learning process. The learning result is therefore not elastic enough to apply in real situations.