Here is an AI feature for you :D => Users share their screens and you detect webpages, and applications on the client side (process frames/images, detect logo, extract text from it, etc.). Don't send the stream to other parties only you share detected apps/webpages with others.
Finally, I've found somebody like me. I was starting afraid about how my brain works. Can I hear the sounds? is that the sound? where are the words? Oh, here I can see now. Hmm, let me try thinking about a class and methods of a module in my codebase... no voice, no words. no images. what the fuck? how I can know that.. yes it's like feeling. Fuck now I'm starting to hear the voice of somebody else. Which was the my voice?