I know I might sound dumb but I had an idea in my head so lemme explain it to you properly. An app that takes your voice input ( EX: Can you show me cake recipes) This voice input is converted into Text format and this Text format is searched in the search engine, which can be done by simple code. Now the result obtained by this search is taken in text format and converted back to voice format and the output is given to the User. Now you might think that this is pretty dumb we can just ask Alexa, Siri to do it much faster and quicker. Yes, you're correct but this App will have Animation involving in it like Talking Tom or just a virtual character having food, sleeping, dancing, doing yoga, etc. Its gonna be like a virtual Character answering the User and the character will have Lifelike Movements, It will properly lip-sync and talk back to the User. Ex: I asked can you dance or sing a song for me then the animation of the character singing a song will be played. I know It's pretty dumb just give me your opinion on it. BTW it's a combination of virtual assistant and fun animation which will be much more compelling than just hearing to Alexa.
I'd find it annoying. Google already takes voice input and to have it spout out the results would do my loaf in. VR is getting better and better and is commonly used for input to a variety of applications.
It's not a bad idea but Google pretty much has it covered with a Google Nest Hub Max - a Google Assistant with a built in screen. Many queries can be answered with Youtube videos.