AutoDraw is a brilliant tool. However, after using it multiple times. I thought of something which can be of help. If I want a flower, I have to make a rough sketch.
I will get this.
Now, if I want a tree. I have to make a rough sketch again. This will require little effort and a few seconds. Imagine if we had google now integration and I could just say “I need a tree”.
AutoDraw still make sense because not all times I know the name of the thing I want to draw. Also if I want the suggestion be close to what I draw.
In a developer’s perspective it makes sense not to do something like this because you miss a lot on training data.
In a user’s perspective this can be of great help.