Post #4280857

@tsyum@thepit.social I think GAN (generative adversarial network) might be broad enough. They (very oversimplified) generate noise and then refine it into something that trips a computer vision model so they "know" they've generated an image. They can be combined with LLMs to work with narrative text. https://en.wikipedia.org/wiki/Generative_adversarial_network