This is one of the many reasons why StableDiffusion is likely to blow away the competition with enough time. The demand for GPIs has grown like people are mining crypto, but because they are trying to train new models of AI image generation.

The open source tools are going to be very hard to outcompete regarding iterative improvements. SD is moving at light speed. Here’s to hoping this turns out to be the case in a world with 100s of thousands of AI models and different systems for every purpose under the sun, rather than vast centralized AI platforms.