Recently, OpenAI announced (with a 99-page Report) the development of GPT-4, a large-scale, multimodal model capable of accepting both image and text inputs and producing text outputs. While the model…