A time you used an AI tool

I'd like to talk about the first time I used an AI chatbot seriously, which was last spring when I had to write a project proposal for a small grant at my university.

I'd used these tools before just for fun, asking silly questions, but this time I had a real deadline and a very messy first draft. So I pasted my draft into the chatbot and asked it to point out unclear sentences and suggest a better structure, rather than rewriting the whole thing for me.

In terms of how well it worked, it was surprisingly helpful but not perfect. It reorganised my ideas into clear sections, background, aims, method and budget, which I hadn't thought about properly. It also spotted that I'd used the word "significant" about eleven times. On the other hand, it invented a statistic about youth unemployment that sounded convincing but turned out to be completely wrong when I checked. That was a real lesson.

My feelings were mixed. I was impressed by how much time it saved me, probably three or four hours. But I also felt slightly uneasy, because I realised how easy it would be to simply trust everything it said. In the end, I treated it like a smart but unreliable classmate: useful for ideas and feedback, but you still have to check the facts and make the final decisions yourself. And yes, we got the grant.