When using the picture mode it would be amazing if it lets you know what it says in any specific language.
For ex. I’m in Japan and I see a sign I can ask rabbit with a picture what the sign says. It would be what I would expect my AI assistant to do.
The R1 already does this using Vision, but the times I’ve used it, it hasn’t worked very well (very poor translation). They should stop using ChatGPT 3.5, which is quite useless compared to the current models.
I noticed that if you take a picture in vision and ask to translate it, it will tell you that translation mode is not possible. instead if you ask it: what is in this picture, or what do you see, it will tell you there is a msg in “x language” and then translate it.
didnt they say gpt4o is integrated?
It works just for translation mode.
exactly! this is one of the things that I would suggest is necessary when thinking about translation and considering rabbit as a assistant when going to other countries.
100 %
to just prompt translate this and take a picture should be super easy and practical.