The future of art has arrived. And it isn’t pretty.
Like, I suspect, many people reading this article, I enjoy the late night comedy shows. When I was a kid, late night comedy was rather dry, filled with stale jokes designed to appeal to the broadest possible demographic. But as we’ve reached the age of peak T.V., so too have we reached the age of peak late night T.V. In our stratified media landscape, where we no longer turn in unison to the same broadcast and news outlets and the idea of a common knowledge base is quickly fading, late night comedians have also taken on the somewhat unfair mantle of often being a source for breaking news. I shouldn’t have to watch John Oliver’s monologue to get a deep dive into real issues affecting the world today. But, with the major networks taking a USA Today approach to most stories these days (only the minimum amount of ink required), late night comedy is often the only place devoting any amount of time to the pressing issues of the day.
This is not meant as an advertisement for late night comedy or a rant against modern news. Rather, it is meant to explain why I found myself watching a story on Last Week Tonight with John Oliver last week about a new service called Midjourney, which allows people to use artificial intelligence to create digital artwork by typing in a series of keywords. While the user themselves isn’t doing any of the drawing or painting, the words they choose to enter result in the computer generating their best approximation of the user’s intent. Predictably, the results range from sublime to awful. And the segment is played mostly for laughs. But the fact that something like this not only could exist, but already does exist, should raise an artist's hackles regardless of discipline.
Only a day or two after watching that segment, I saw a news story about an A.I. rapper that lost its record deal. Yes, you read that right. And I have so many questions. One, how did a rapper that literally doesn’t exist get an actual record deal? Two, how did they teach the A.I. to rap? Three, who is behind the A.I. rapper? As it turns out, it’s that last part that led the rapper’s contract to be rescinded and leads to more pressing questions of why it was offered a contract in the first place. As it turns out, the digital gangster rapper, decked out in every stereotype of what certain people seem to think of African-Americans in general, was the brainchild of two non-African Americans. I’m not going to go into the entire history of blackface and minstrel shows in this essay. That is a topic for its own essay, book, and/or documentary series. But, let’s just agree that a digital minstrel show is no better than a live one.
Of course, the echos of history didn’t stop there. The voice of the A.I. avatar who secured the record deal was actually that of a real rapper. But, much like many a shady music producer throughout the history of the industry, the producers had manipulated the situation to take the talent from the human artist, convert it into more easily manipulated ones and zeros, capitalize financially on the artist’s work, then cut them out of the financial profits all together.
As more facts about this A.I. rapper came to light and more outrage began to build, the record company canceled the contract and scrapped their plans. But why could they not have seen the problem with the plan upfront?
And, stepping away from any cultural appropriation issues for a moment, what implications does that have for the music industry going forward? We already live in a world of autotune which can make almost any average shower singer into a songbird. As someone who is “musically challenged,” I appreciate the idea that a computer can make me sound like Marvin Gaye. But does that mean that I can actually sing? I’d say no. But what the story of the canceled rapper tells us is that we’ve rapidly reached a point where the human being might not be needed at all. Just plug in a handful of appropriate hashtags and voice samples and voila! You’ve got the next Whitney Houston. And, better yet, you don’t have to pay her a cent.
The same thing is coming to the visual arts as well. I was watching Trevor Noah’s show last night (yes, another late night comedian), and he had a story about an A.I. artist who won an art contest. The winning image was 90 percent created by artificial intelligence, according to the artist. He submitted the original keywords and did some finishing work in Photoshop. But the bulk of the heavy lifting was done by an algorithm. Without commenting at all on the quality of the result, what larger questions does that pose? Is the submitter really the artist behind the image at all if he only contributed 10% of the work? He did come up with the original set of keywords to generate the image. So, it’s fair to say the image wouldn’t exist without him. But all the intricate brushwork, lighting, and composition that can take a human being decades to perfect were all done by the computer. I don’t know if his winning image would technically be termed a painting or a photograph. It’s very photorealistic. But, even if it’s technically a painting, the implications for photographers are obvious.
Why would a client pay you thousands of dollars in creative fees and licensing to take a picture of their products when they can have a computer generate an image for them? True, it probably won’t be totally free. No doubt, a cottage industry will sprout up around A.I. images similar to creating VFX for film and television, and there will be new leaders in the emerging field demanding higher rates. That’s how capitalism works. But does that mean that we are on a precipice of seeing our entire professions replaced by machines?
Let’s take the example of a product photographer, for instance. Let’s say Coca Cola needs to make a new ad to promote the introduction of their new flavor of Coke. The ad is going to be a spinning soda can, with music and text. It is entirely possible that it would be far more cost effective for Coca Cola to type the dimensions and hex values of its can into a computer and have it generate the photorealistic imagery of the spinning can than it would be to hire a full video crew. As mentioned earlier, they can even turn to an A.I. musical artist to generate a jingle for the spot that won’t require them to pay royalties.
As someone whose work generally centers around human beings rather than products, I could easily delude myself into thinking the computer couldn’t possibly do what I do. But that would clearly be presumptuous on my part. Assuming that others share my same aversion for ones and zeros. I mean, honestly, even if you’re a fan of Marvel movies, what are they really if not just 2.5-hour tributes to special effects. And, based on ticket sales, people don’t seem to mind.
We’ve been debating ethical issues surrounding digitally manipulating human performances since back when Ridley Scott was forced to use a CGI version of actor Oliver Reed for parts of the film Gladiator following the actor’s sudden death three weeks before the end of principal photography. In the ensuing 22 years, the technology has only gotten better to the point where the majority of blockbuster movies these days contain large amounts of avatar stuntmen and women doing feats of strength that defy human ability. I won’t even get started on the absolute mediocrity of the modern superhero film that seems to not only utilize A.I. avatars on screen, but apparently uses A.I. algorithms to write the scripts as well. Okay, I’m being a little snarky there. But, honestly, are you truly sure that Netflix doesn’t have some kind of reverse algorithm that knows exactly what movies you are watching and directly feeds this information into another A.I. machine that is writing scripts designed to hit the broadest common denominator?
So, it’s not even remotely out of the question that in a society which increasingly views art as mere content, that sheer practicality will dictate that the majority of content we consume within 10 to 15 years, if not sooner, will be created almost entirely in a computer. The question is, as artists, what do we do about it?
The enterprising capitalist may look at that scenario and decide the only way to make money as an artist in the future is to become the puppet master behind the creation of A.I. art. If I were Benjamin Braddock’s plastic loving neighbor in 2022, it would probably be A.I. that I would be urging the hapless hero to get into. But as an artist who straddles both the analog and digital generations, the idea of giving up even an inkling of my artistic output to a machine just somehow seems wrong. That’s not to say that it is wrong. We all have our own levels of comfort with how much of our art should be computer generated and how much should be done “in camera” so to speak. I personally tend to reject computer generated art instinctively, like a body rejecting a blood transfusion. But that’s entirely subjective, no doubt affected by my age and personal preferences. And, as CGI gets better and better, the line between computer generated art and the real world continues to blur. So my own red line will undoubtedly shift.
The other day, I was rewatching one of my favorite movies, Braveheart. The film is famous for its massive battle scenes with hoards of opposing armies going toe to toe across sweeping vistas. I didn’t know it when I first saw it in the theater, but in order to have that many soldiers in the battle scenes, a large number of the soldiers were created digitally. This is commonplace now, but at the time was revolutionary. They did have a significant number of extras to play soldiers. But they essentially doubled and tripled those extras to fill out the scene. To me, this feels like a good use of digital technology. The filmmakers still accomplished the bulk of the work practically. They simply rounded out the hard work they had already done with these digital soldiers so that they wouldn’t break the budget. I never noticed it in the film, and it didn’t bother me.
Compare that with the modern superhero movie where not only the actors, but the environment, and half of the props are created whole cloth from digital assets. The actors give performances against green screens (or the new, more immersive virtual LED walls), and everything around them is literally created by a computer. There are still, at this moment, real human beings serving as VFX artists to create those digital worlds. So, it’s not the same as an A.I. generated environment. The human VFX artists are truly digital gods. But I do find myself emotionally disengaging from the majority of these modern action films simply because it’s impossible for my mind to divorce from the fact that I’m watching one and zeros turned into avatars rather than actual people.
People, as imperfect as we are, have value. In fact, it’s our flaws that give us that little extra something that allows an audience to relate to a character and see a little something of themselves. Human beings have oddities that make them special. Audrey Hepburn complained about having an extra long neck, but I defy anyone to see her as anything less than beautiful. Humphrey Bogart had a lisp. Jimmy Stewart’s awkward and stilting way of talking not only didn’t hurt his career, but became one of his trademarks.
A few months ago, I was revisiting John Woo’s 1997 action film Face/Off. There’s a scene in the film where the two characters, played by Nicolas Cage and John Travolta, are in a speedboat race. At some point, there’s a boat crash and the two characters are thrown from the boat, flying high through the air. As many times as I’ve seen the film this particular time, it became painfully clear to me that the people flying out of the boat in the wide shot were quite decidedly not Nicolas Cage and John Travolta. This makes sense. This is why the profession of stuntman exists. And it’s not like, deep in my brain, I didn’t already realize that there’s no way the two stars would do that stunt themselves. But, in the context of the film, I had suspended disbelief long enough to just go with it. And, you know what? It totally worked.
Nowadays, those stuntmen would have been replaced by digital avatars. They would have been able to superimpose Cage and Travolta’s faces onto those avatars and likely created the boats and the flailing bodies with VFX. But, somehow, I seriously doubt it would have felt as exciting. There’s something about seeing a real person, even an imperfect one, in peril. One of the main reasons the new Top Gun: Maverick sequel works so well is because of Tom Cruise’s dedication to practical stunts. Surely, the movie has its fair share of digital effects. But by keeping those to a minimum, you allow your audience a greater ability to connect to the story. Because what you are seeing is real. It’s human. It’s relatable.
So, as we continue to hurtle our way towards a world where A.I. is going to start taking away a larger and larger chunk of the creative jobs in which many of us make our living, I suspect that the most powerful tool we will have in our defense is our very imperfection. This sounds counterintuitive, but what’s the one thing about us that a computer will never be able to replicate? Our humanity.
If your entire artistic voice is based on your ability to expose perfectly to your light meter, you may find yourself in trouble. Computers can already perform that trick. Just think about all the tasks your camera is already capable of doing for you. But your camera, no matter how many megapixels, can’t replicate your artistic voice. Your artistic voice is the culmination of all the experiences that you’ve had in life, art-related or otherwise. Your artistic voice is the grand sum of your life's emotions. It is when you put yourself, all of yourself, into your work and art that you can create greatness.
There will no doubt be more and more pressure to turn over more and more of your artistic creation to technology as time goes on. And, no doubt, new technologies will come along that will genuinely afford you a chance to be a more effective storyteller and artist. But ask yourself why you wanted to be an artist in the first place. Was it just to get a result as quickly and as cheaply as possible? Was it just as a shortcut to fame and fortune? Or, did you become an artist because you had something to say? Did you look at the great ballad of life and feel that you had a right to contribute a verse?
There’s no question that the machines are coming for us as artists. But trends and technologies can never replace the value of human intuition. Our creativity will sustain us. Our humanity will sustain us. And we, as artists, will survive. No matter what SkyNET says.
Years ago there was a discussion about computer vs chess player. Conclusion was that computer will never be able to beat a grand master. Well, years later computer was able to beat world champion. Just today I signed up to site called Graiphics (you can google for Youtube reviews) for $35. Just for kicks I asked it to generate a picture of Bear lake from Rocky Mountain National park. I was shocked to see the result. It resembled the picture I took two years ago from about the same viewpoint. Then I asked it to do the same in winter and it did it! Granted, Bear lake is one of most photographed locations in Colorado, but if I needed a picture that was unique and didn't violate any copyright this AI generated picture would do. Of course I could get stock picture for very little money, but it would be a same as one on hundreds of other websites. My point is that artistic voice has very little value when it comes to commercial use of the photography. And I'm sure that in not so distant future AI will be able to replace 99% of so called artistic photography. How long will it take to pro photographer to photoshop astronaut riding horse on the moon? It takes AI minutes to do that.
But you don't know that it was "unique and didn't violate any copyright".
I don't know if this particular software generates truly unique image. They say it does and explain how AI creates it. But there is no doubt in my mind that there will be many startups that will improve the software to the point that AI created images will be at acceptable quality for many publications and websites.
There is currently no copyright protection on any AI imagery. Which means, that once it's generated, literally anyone can use it as they please. So if you create something using AI, there's nothing stopping anyone from taking it, printing it on shirts and making money from it. And there's nothing you can do about it. Unless the US Copyright laws change (which they could), any company using AI imagery to promote and sell their brand risks competitors and other companies using those same exact images. No different from stock imagery.
Or this competitor could just use the same software to generate their own image similar to yours for very little money.
Yeah I suppose that could happen. Depending on how closely it is to copying my work however, it could become a copyright infringement of my work. No different than an artist recreating a Van Gogh painting would be a copyright infringement if it were to be monetized in any way.
Artists should fund the creation of an AI CEO and see how quickly the use of AI "workers" gets frowned upon.
Really well written article Christopher Malcolm ! There are a lot of valid points in here. I don't really like where a lot of this is headed. It's hard enough as it is to make a living as an artist and I can't help but feel like AI is on a path to seriously destroying and devaluing the work that we do. There's a lot of uncertainty at this point in time.
James Earl Jones recently gave the blessing for a company to use AI to reproduce his voice for Darth Vader: https://techcrunch.com/2022/09/26/ai-is-taking-over-the-iconic-voice-of-...
I've seen songs popping up on youtube that were created completely by AI that are writing "Nirvana" songs. How long will it be before we start hearing music on the radio and on streaming services that are completely generated by AI? I didn't even know about the AI rapper that you mentioned in the article.
I think nearly every industry can and probably will be affected by AI at some point or another.
I saw that about James Earl Jones. I was like "noooooooo." Then again, I completely understand him wanting to cash out. I'm sure whatever they are paying will take care of many generations to come. But, once they can create James Earl Jones with a computer, it's a slippery slope.
Yeah it's pretty crazy. I see this being a trend with celebrity voice overs in the future. Bad news if you're a celebrity voice impersonator I suppose. Only time will tell