As artificial intelligence continues to be improved upon, it will become better than humans at many tasks.  From driving, to retrieving medical information, it appears that there are an incredible number of ways AI can be utilized as outlined by Stanfords One Hundred Year Study on Artificial Intelligence.  However, there was one claim that artificial intelligence has the ability to create art and music.   I refute this point by claiming that art can not be created by artificial intelligence alone, for it lacks the ability to be creative and express itself.  
	 Humans are the ones responsible for creating the methodology for programs to reach a conclusion, not an expression.  In order to explore this idea, Googles deep dream generator can be examined.  Arguably the most iconic example of visual manipulation through the implementation of AI technology, this program has the ability to take two images and create a mash-up using deep neural networks.  The program creates a psychedelic transformative picture that was put together completely by AI.  In addition, Google has started a project called Magenta that has the ability to generate pictures and musical compositions.  Even though AI is able to produce images and compositions, the actual product is not art created by AI.  As futurist Jaron Lanier explains, AI assisted media ultimately still comes from people, and the problem from that is that the people are made anonymous.  This means that media created with the assistance of AI is technically art that can be attributed to the person that contributed the newest input in addition to the humans that coded the program.
	The article goes on later to claim that in the future, AI will simply help humans create media.  Interestingly, this portion of the article was incredibly vague in how AI would be able to assist in creating media.  It is important to make the distinction that true art must be the result of some human effort, for it is still up to humans to make the creative decisions.  Why does this matter?  This exposes a limitation of AI.  Programs produce a result determined by directed paths based on predetermined logic.  Human expression is not always based on logic, but often based on emotion.   
	That leads to one final question: will AI ever be able to express emotion?  The answer to this is complicated, but the answer is that it is unlikely since there are many physical and natural reasons why humans feel emotions.  Many emotions can be informed and altered by factors such as the air we breathe or the foods we eat.  Fear, disgust, and nausea are all created by external factors that would be almost impossible to emulate.  In addition, love and sex are based on humans innate need to recreate.  Finally, robots can not mature in the way humans can.  AI lacks the ability to emotionally grow, experience hormonal changes, or have other chemicals influence thought.  
	In conclusion, AI needs to be able to express emotion in order to truly create art.  Without the ability to inhabit a human body, AI will forever be limited to being a tool in order to help humans create media.  Despite this, it will be fun to see what artists and programmers can produce in the future with the help of new AI technologies.  

Sources: http://hplusmagazine.com/2014/04/29/could-a-machine-or-an-ai-ever-feel-human-like-emotions/

http://www.popsci.com/can-computer-make-art


