What industries and problems could benefit from GPT4 accepting image input and providing text output

Hello Product Hunters, I'm curious to know which industries and real-world problems you think would benefit the most from the integration of image input and text output capabilities in GPT-4. Your insights and perspectives are highly appreciated. Thank you!