I'm super curious what the ML looks like behind the scenes. Do y'all think it's just a GAN thats been fed tons of logos that already exist? That seems unlikely–I can't imagine there are nearly enough logos for that.
Ha most people are. Unfortunately, it's not a GAN that does everything. We wanted a system that works and creates high quality logos and GANs are still reaching there. On a high level, there are multiple ML models that are responsible for different tasks such as color combination generation, icon selection, font pairing etc. All these collectively form the entire product. Let me know if you want to chat more about this.
