Before jumping into how, let's quickly discuss the why in order to motivate us as well as encourage creative exploration of this concept. As alluded to in the introduction, the first motivator is to reduce friction. Consider the soft keyboard (keyboard with no physical buttons) on your phone; due to the constraints of the medium, such as lack of space and feedback, inputting text without predictive text would be cumbersome to the point of rendering it unusable. Similarly, despite the convenience of drawing with our fingers, our fingers are not that accurate, which makes things difficult to draw.

The other reason why this concept (augmentation) is advantageous is its ability to democratize the technical skill of drawing. It's common for people to not even attempt to draw because they have convinced themselves that it is beyond their abilities, or possibly that we can enhance one's ability to draw. This was the motivation behind the research project ShadowDraw presented at SIGGRAPH in 2011 by Yong Jae Lee, Larry Zitnick, and Michael Cohen. Their project had shown that guiding the user with a shadow image underlying the user's stroke significantly improved the quality of the output.

Finally, the last reason I want to highlight as to why this concept is interesting is providing a way for users to work at a higher level of abstraction. For example, imagine you were tasked with sketching out the storyboard for a new animation. As you sketch out your scene, the system would substitute your sketches with their associated characters and props as they were being worked on, allowing you to design at a higher level of fidelity without sacrificing speed.

Hopefully, by now, I have convinced you of the potential opportunity of integrating artificial intelligence into the user interface. Let's shift our focus to the how, which we will begin to do in the next section.