Currently, as someone moving into college and trying to pursue his educational career, I often find myself using chat-GPT as a tutor or even a counselor at times. One thing that I find inconvenient is that every time you want to answer a question or make a request, you have to manually put it in. I do sound lazy but my idea is a web extension or application that works as a more hands-on tutor/guide. Currently, I’m envisioning an application that can further enhance chat-gpt’s image recognition by using it through various tasks. For example, using the already readily available image recognition chat-GPT offers can guide me WHILE taking multiple-choice questions or homework. Instead of swapping tabs, it can guide me through the whole process as I’m doing it. So going back to my example, if I were to choose the wrong answer, it can guide me with physical notes on the side or even arrows to point me in the right direction.
Another great example is maybe doing math homework. For example, while doing math homework on my iPad or computer, it can point arrows at places that I messed up and help correct me instead of having to input all my work into the chat-gpt’s response bar.
This is more so directed towards students who are struggling in school and cannot/do not have the resources to compensate. Maybe you have a bad teacher or maybe just struggling with a certain topic. I see this idea as not far out of the reach of engineers and researchers at Chat-GPT. You already have to resources to make it happen. All in all, this idea I can see being useful in any aspect of learning not only in a school setting.