Evolving Our UI
When thinking about how to design an AI-powered sketching app, "input" was an area we spent a lot of time thinking through. We knew we wanted to limit the number of input options because this was not just a regular sketching app, and the system for it to work had to have context.
Therefore, limiting the initial input to Text / Voice afforded us a better interaction model, especially since it was a natural call-to-action because of how the system associates images through NLP language processing. It was also a delightful moment we found to inform the user there are in an "enlightened" experience.