It baffles me people still care about code quality more than interaction speed. 

Each millisecond affects the user experience.

building far.quest & @wieldlabs.eth. cofounder at miy.com ventures. FID #274. Previously Square/Cash App. https://hiimnico.xyz/

I'm curious if users are willing to make the tradeoff long term with LLM powered products, especially with how slow token generation is.

It's not perfect, but adding a visual component to the "thought process" of agents we've built has made the wait a lot more understandable. 

For complex problems, there's inevitably going to be a delay in processing due to the natural latency of working with multiple systems. Maybe adding tips and tricks like games?

have you seen any good UX for LLM powered product, besides token streaming?