Open AI has been focusing on the wrong things
stupid chatbots
OTOH, they demo'd something that I really want... instant two way translation of spoken language... but to date they haven't released this yet. Probably because of the enormous amount of resource the AI takes.
I want to have a conversation with someone who either doesn't speak my language or doesn't speak it very well. Whey they speak to me, I want to hear English translation AS THEY SPEAK, then I want to reply and have my reply instantly translated to their language.
Bonus points if you can do the "babelfish" thing ( HHGTTG ) with "brain waves" instead of sound ( i.e. do the translation job, convert the tone to my language but leave it in the voice of the person talking and stimulate my vestibulo-cochlear nerve ). That way, I can listen to a conversation with myself and a few others all speaking their native language ( my voice can come from a device like my cell phone ).