Do you have a dumb question that you're kind of embarrassed to ask in the main thread? Is there something you're just not sure about?
This is your opportunity to ask questions. No question too simple or too silly.
Culture war topics are accepted, and proposals for a better intro post are appreciated.
Jump in the discussion.
No email address required.
Notes -
During my drive into work the other day, I realized that I could have a voice conversation with ChatGPT, to keep my commute productive and explore some ideas in the morning. It worked out for some simple back-and-forth but as soon as I tried lengthier conversation, the flow just wasn't quite there. ChatGPT thought that pauses between sentences were the end of my prompt, and would start responding even though I was still thinking about what to say. Voice mode currently feels like a glorified text-to-speech engine with shorter responses.
Note that I'm not belittling the effort it has taken for OpenAI to release voice mode, but its capacity for natural conversation is still far away, though given the advances the AI space has made in the past year, it's probably sooner than we think. (You can even dial 1-800-ChatGPT and it'll allow you to use voice mode without the app!)
Has anyone here had any success with LLM services that are tailored for natural conversation?
A quick hack you can use is to instruct it to use "radio etiquette", and not to respond with a full answer until you finish a transmission by saying 'over'.
The way it is coded, it will be compelled to respond every time you "send" it a prompt (which happens automatically when you pause), but you can instruct it to only respond with e.g. a single '.' (which it will read, but that's only one syllable) until you say 'over'.
Funnily enough, it's role playing conditioning will automatically also result in it saying 'over' after every answer.
More options
Context Copy link
More options
Context Copy link