#Still speaking notifier
1 messages · Page 1 of 1 (latest)
That would require every other collab partner to install something (any decently aware person wouldn't do that immediately), or watch a screenshare of Neuro that is separate to the stream so it doesn't say on stream that Neuro is still talking
but yeah there should be some speaking notifier somehow if possible
I wonder if there’s something you can do with discord.
Like on the voice channel where neuro speaks maybe have a livestream that just shows a green or red screen that indicates if neuro is talking
probably just some visual cue next to their model would do, something similar to the ? they have now
actually, now that I think about it
there is a push-to-talk thing on discord UI, if there is a bot version of that, vedal can program it to use push-to-talk and just hold it until like idk 1 full second after she finishes talking? (it outlines the user using it with the usual green line, even if there is no audio input)
Personally I feel like that would make the conversations with them less natural? In human interactions we speak over each other unintentionally all the time. It might make the collab partner feel more like they're talking to a chatbot rather than an ai with personality. Plus it will be one more thing for them to pay attention to. That's just my opinion though it might be worth a try.
I would agree with at least experimenting with it first
From what I see it usually takes just a few minutes and the collab partners will naturally find the correct rythm
Also with a notifier it becomes less a natural conversation and more of a genuine chatbot experience
what if collab partners could see their control panel with the full message? Like what vedal can see during streams?
though I suppose a more "natural" solution is to make the twins much harder to interrupt mid speech. Partners should still be able to interrupt if they want to, but things like laughing or other interjections shouldn't make them cancel their speech
or here's a crazy idea, maybe they should just give shorter messages in general? I know we all love neuro's crazy long rants but in real conversations people don't prepare multiple sentences in advance before they start speaking. If they can split sentences into shorter sections which they can adapt on the fly as the context changes maybe it could be way more natural? (I have no idea honestly)