News for hackers. Stuff that matters.

A comment on We're expanding our Gemini 2.5 family of models. Parent comment

Re:We're expanding our Gemini 2.5 family of models by jjani on Tuesday June 17, 2025 @07:22PM (#44305051)
Correct, though pretty much anything end-user facing is latency-sensitive, voice is a tiny percentage. No one likes waiting, the involvement of an LLM doesn't change this from a user PoV.
[ Parent | on HN ]
Re:We're expanding our Gemini 2.5 family of models by eru on Tuesday June 17, 2025 @09:21PM (#44305775)
I wonder if you can hide the latency, especially for voice?

What I have in mind is to start the voice response with a non-thinking model, say a sentence or two in a fraction of a second. That will take the voice model a few seconds to read out. In that time, you use a thinking model to start working on the next part of the response?

In a sense, very similar to how everyone knows to stall in an interview by starting with 'this is a very good question...', and using that time to think some more.

[ Parent | on HN ]