Or are critics safe from the takeover?
Some say the robot takeover is already among us. Companies like McDonalds and Panera Bread have already taken to hiring kiosks in place of regular cashiers to make the entire ordering process less confrontational for everyone. Amazon on the other hand has gone even further down the line, already using their Kiva robots to sort merchandise for order so that us humans can rest our legs- albeit perhaps in lieu of a paycheck.
“‘Robots are essential for meeting that kind of demand,’ said Ken Goldberg, a robotics professor at the University of California, Berkeley. Humans just can’t work as fast.”
That’s where the worry begins to come in, and YouTube education channel CGP Grey seems to have the scary possible reality figured out in their 15 minute-long piece Humans Need Not Apply.
Have you ever felt comfort in thinking “A robot could never do MY job”? You might feel a little different after you watch this.
The video features takes on several different components of everyday American life and employment from automobiles and white collar work to the special creative jobs that are generally believed to be safe from the robot takeover. The narrator proceeds to pick apart each and every component and explain how robots can not only handle them, but do so better and more cheaply.
And the aforementioned “special creative jobs”- yeah, those aren’t safe either.
“The brain is a complicated machine- perhaps the most complicated machine in the whole universe. But that hasn’t stopped us from trying to simulate it.”
The video goes on to explain how replication of the human brain for creative work has already occurred in the form of music. There already exists Emily Howell, a computer program created by UC Santa Cruz music professor David Cope. Emily can generate completely new music using a source database and can even “listen” to human feedback to further expand on her compositions.
Feel free to give her a listen below. Emily could feasibly have another beautiful song composed before you even finish this one.
So after all this, what about music criticism? There clearly cannot be a ray of hope for the many people out there making their livings writing album and track reviews. Could robots really review music someday?
Probably not. Consider the following.
If we took the imperfect human mind and replicated it to create the perfect version of the “most complicated machine in the whole universe”, think of the music critic that thing could be. A robot with this “mechanical mind” could simply scan an album’s audio once and quantitatively measure all the variables of what makes a good album: everything from instrumental variety and production to lyrics and themes. It would essentially be a perfect listening experience.
So why “probably not”? The answer lies in the distinction between listening experience of a robot and that of a human: the robot’s experience is perfect, while the human’s is fundamentally flawed in comparison. It just so happens however, that a “flawed listening experience” might be key to what makes a great review.
Here’s an example.
“Micheline used to come to our house and knock on our door”
The album Benji from indie folk act Sun Kil Moon was my favorite album of 2014. There was no argument about it. The album’s incredibly descriptive storytelling and ability to let in rays of hope through the cloudy themes of death and tragedy are inspiring. You probably wouldn’t ever guess, then, that I hated this album at first listen. And do not even get me started on the second listen- it was absolute torture listening to this guy ramble on about random deaths in his family while plucking away at his acoustic guitar.
My listening experience was flawed because it took me several listens to absorb enough of the record to the point at which I could develop something even close to my final opinion and write a proper review. Hell, I even made a complete 180 by the time I finished the third listen. A robot should honestly just write the review instead, right?
Well, the robot might listen to the same album one time and gather all the data there possibly is to connect to make the quantitatively best possible critique of the record. This said, what makes a review great is the human listening experience- perhaps even how the Benji grew more enjoyable with repeated listens.
Our “flawed listening experience” is what gives our reviews humanity.
After all, people read music reviews primarily to compare their listening experience with that of another human being. They seek a human connection in that way that simply cannot be established between a human reader and a robot reviewer. Sure, the robot may have the objectively better and more accurate review. But incredible quantitative accuracy is not the defining characteristic in an album review.
As far as most readers are likely concerned, relatability is far more important. And who can a human relate to more than another human?
Music criticism aside, there very likely exist all sorts of minor issues with major consequences as far as robots taking over for jobs all across the board goes. My argument is one of many, and I encourage you to think of what issues might arise if artificial intelligence takes your paycheck sometime soon.
In the meantime however, my girl Emily Howell is currently composing an unlimited amount of music, so it looks like I have some listening to do.
image via Carnegie Mellon University