AI and human responsibility

One night, while building EduVision, I couldn't sleep. It wasn't a bug. It was a thought that had just occurred to me: if my AI reads a word wrong, and a student trusts it, whose fault is that?

In an ordinary app, one wrong word is just an annoyance. But for someone who has to trust your voice completely, a wrong word can lead them somewhere wrong. I realized that when you build for vulnerable people, no small mistake stays small.

I'm still a student. I don't have answers to the big questions about AI. But I believe responsibility starts somewhere very close: check a little more carefully, be honest about where your tool is still weak, and never promise more than it can do.

A lot of people like to talk about how AI will change the world. I care about a smaller question: when my tool touches a real person, was I careful enough? Until I can answer that, I don't dare call myself someone who builds technology decently.