Do you think AI is a threat to artists? 
Below are a few AI-generated portraits of myself. They actually look like me. I paid $10 for them and it took 20 mins to generate 100 portraits!
I’m fascinated by AI art and was reading about how entry-level illustrators are slowly being replaced by this technology. Most popular websites use stable diffusion and only require an input phrase to generate an image. For example, I could type, “monkey wearing a helmet in space in hyperrealistic style” and out the other end, you’ll get something resembling that… depending on the parameters set.
The first time I played around with AI art/generated images was back in 2015 with Google’s deep dream software. It was initially designed to search for patterns in images to classify them. However, the process can be run in reverse… for example, by finding dogs in an image. It’s been compared to staring at clouds and finding animals. It overprocesses an image, finds patterns, then repeats this over and over.
AI-generated images are only becoming better and more convincing over time. Some of them look so good, I couldn’t create them myself!
I see potential in using AI as an assistant to come up with ideas. Sometimes I search around for new reference photos etc and base my art loosely around those. Experimented the other day by uploading some of my own images and adding a phrase – you can get some interesting results. There’s also a degree of human involvement. You can select the ‘best’ of say 10 generated images and further process that idea. All the initial prompts and parameters such as how much the program adheres to your prompt can be set. Anyway, what do you think? 
