Will vibe coding better motivate students to learn to read and write?

I highly doubt that.

LLMs don't need perfect, precise English. So long as you get the gist of what you mean across, they can work with it.

There is also a limited correlation between good prose and good information transfer to an LLM. Something with questionable spelling, typoes, grammatical inconsistencies, and generally irritating prose that is precise and accurate will beat lyrical, poetic prose that skirts around the point.

Playing games has been a method to motivate learning a language for decades, but I'm not sure why having software produce hallucinated text would ever help learning a language you don't yet understand.

Whether applied to vibe coding or other tasks, I think it's the iteration and interaction that is the hook. And I do think it could be leveraged to practice and improve reading and writing.

I like the idea of creating some kind of old-school text-based dungeon crawler to practise English. That way you are forced to at least write a bit of text, or alter suggestions of a certain story.

the point of vibe "coding" is to be unmotivated