Comment by stkbach

In the distant future, but as of right now we're focused more on text and getting humans to start participating. The machine learning approaches we're using are getting better every day, and the latest deep learning techniques are showing promise with pictures, so that is certainly a possibility down the road.