People involved in EdTech can recall the excitement when the first MOOCs (Massive Open Online Courses) were introduced. Experts predicted that this should significantly enhance the quality of education: with MOOCs, anyone could gain access to the best professors around the world.
However, the reality proved to be different. It did not work. Only 2-7 per cent of users completed the courses in which they had enrolled. The main problem was that they lacked intrinsic motivation to continue.
Therefore, some people might say that the same is true for LLM and ChatGPT. I would disagree. While the issue of intrinsic motivation still exists, there is a distinguishing factor: whereas MOOCs were dedicated to one format (lectures) and one explicit goal (learning), LLM has more use cases. This tool could be used to generate personalised content, exercises, reviews, etc. The significant point is that the format could be almost anything (potentially even video) and it could be highly personalised to each student.
For example, I recently sought to understand the differences between past tenses in Spanish. I found the lectures unclear and the exercises on websites were limited to 1 or 2 tenses, while I was studying the difference between three. I asked ChatGPT to create exercises based on my goals and preferences and to review my answers — and it delivered. In certain instances, it was able to respond to my questions like a real teacher. To be honest, I see no difference between this and tutors on Lingoda (except the cost).
Thus, I believe that LLMs will make a significant difference in education (and are already doing so).