I feel you have misunderstood this post. By saying that we are training a machine, it by no means implied that it understands, has cognitive capabilities, or are sentient in any sense. It is a means by which we can try to recreate the repeated patterns of certain human mental process, and re-apply it in large volumes.

Our innate motivation of building such technology is always as an aid to human creativity, not a replacement for it. For example, pain of organizing and curating photographs is real and getting exponentially worse every day, and we need smart tools to solve it. What the technology gives is a start point to organize and order one’s content. But the user is in full authorial control to over-ride its decision.

Photography ( and consequent visual mediums like cinema ), started as a technology, but artists quickly found it as means to convey their expression. The interaction between creativity and technology has always been an interesting dynamics. So it will be fascinating to see how it will evolve in next few years, since technological space is going through an acceleration phase right now, a big part being powered by advancements in machine learning.