Hi @lukasmoeller,
I saw your work around Cursor’s inference engine and wanted to ask a genuine product question.
We’re building in code generation at Kog, and one thing I keep wondering is where the UX really changes once generation becomes dramatically faster, not just a bit better.
At what point do you feel speed starts unlocking genuinely new product behavior rather than simply making existing flows feel smoother?
I’d be very curious to hear how you think about that.