We know a lot of people already use it. We won’t be very original by agreeing that it can both be of enormous help, but that it bears a fair amount of risks.

At Kollin, we have decided to integrate it because it can have a lot of potential to help students solving exercises. It may not be always true, but students are generally well aware of that as digital natives. While not all answers are true, they can at least provide a good food for thought and give ideas about how to think around or what to do.

Together, students can help each other out with ChatGPT and find good prompts, review its answers, view other’s conversations with it. Instead of using it in silos, and having to reinvent the wheel, in terms of prompting. This actually fits well in our idea of collective intelligence. Students provide a lot of valuable work, too often in silos, and we see a space where they can share their efforts, to build a network of knowledge. We think students deserve a good study experience that matches their expectations. We don’t pretend we are there yet, but this is what we are intending to build. Collective intelligence is what brought a lot of things to life (for example Google Maps). We think AI is a great complement for this network, to enhance the experience.

What we find especially interesting in ChatGPT is that the richer the context, the better the answers. So we can help providing a good context, for example the text of an exercise.