People probably have problems that can be solved using a Telegram bot. The question is hard to answer though, because we might not consider we need a Telegram bot to solve those. Generally it's more productive to start with the problem, and choose the solution/technology based on that.

What makes you ask this question?

Good point Marc. I brought it up because I was starting to play around with the API and thought it'd be fun to tackle a real problem somebody else was working on.