GANs have already proved useful for image generation tasks. Most of the research in GANs is currently focused on high-resolution image generation, text-to-image synthesis, style transfer, image-to-image translation, and other similar tasks. There is not as much research at the moment into using GANs for text generation. This is because GANs were designed to generate continuous values, so training GANs for discrete values is really challenging. In the future, it is predicted that more research will be undertaken in text generation tasks.