Will AI make artists obsolete?



Two years ago, Prisma came out — an app that lets users turn their photos into works of art.

Other apps have hit the market since then, for all platforms, including mobile, desktops, and Web browsers, offering various degrees of customization options and image quality.

Some apps are free, or have free basic versions.

Now that any picture can be turned into a famous painting, are artists obsolete?

The short answer is: No.

Here’s why.

Barriers to art have always been dropping

The new AI-powered tools put a lot of power into the hands of users, regardless of their level of artistic skill.

But every new generation gets its new set of tools. I bet when commercial paints first became available, older artists complained that it wasn’t real art unless you went out into the forests and gathered your own ingredients to make the inks.

And then photography was supposed to put an end to artists. After all, what’s the point of hiring an artist to paint your portrait — which takes years of training — when you could hire a photographer to take a picture, instead. And a photograph is much more realistic than a hand-painted portrait could be.

But original paintings are still worth money. Sure, there are starving artists out there — and there always have been — but original art is still valued.

The fact that art looks easy doesn’t mean its worthless

How many people walk into museums, stare at million-dollar paintings, and say, “My five year old could draw this”?

That doesn’t mean that the artist knocked off the painting in five minutes.

It doesn’t mean that they didn’t, either. Maybe it did only take a few seconds to throw that bucket of paint at the canvas, and the rest of the time was spent cleaning up the mess.

Or maybe it took the artist hundreds of tries to get a result that conveyed what they were looking for.

Digital art is much the same.

It could be quick. Or it could take days, weeks, months, or even years to get an image looking the way the artist wants it to.

Commodity art is a commodity

You might think that Taylor Swift’s music could just as easily be generated by a computer. And you might be right.

But a computer that generates hundreds of Taylor Swift-style songs isn’t going to have her music career. The best it’s going to do is be elevator music, or commodity background music — the stock photography of the audio world.

People buy her music not just because it’s catchy, but because it tells a story about a person we care about. It expresses her emotions, and, by doing so, makes us feel them as well.

By knowing the intent behind the music, the story that goes with it, the personality that created it, we value the music more, we appreciate it — and we pay for it.

Similarly, great art is just as often the story of the painters, or the subjects, or the process of the art as it is about the product itself.

There are plenty of forgeries of the Mona Lisa out there indistinguishable from the real thing expect through expert analysis. That doesn’t reduce the value of the original Mona Lisa.

You could get a free copy of the painting to use as your desktop wallpaper, or a $2 poster to hang on your wall. From a distance, and with the right frame and lighting, most people wouldn’t know the difference. Except of course we’d know. The original is in the Louvre. Everything else is just a copy.

The original creation, the real thing, has meaning and value because of the emotional context, because of the story behind it, because we care.

Art expands to fill the time available

However much time an artist has to create, that’s how much time the art will take.

AI-powered tools may speed up certain parts of the process, but that just means that the artist will spend more time elsewhere.

After I drew a line picture of a cat in a box, it only took seconds to run it through the Deep Dream generator.

But then I ran it through again, with different settings, and different filters. Then through other AI tools, then edited it with image editing software.

In the end, I ended up with more than 50 different final images, only one of which I really liked. Only one that felt to me that it was getting at what I was trying to convey with my cat in a box.

Who owns the AI art?

If you throw a bucket of paint on a canvas, does the bucket own the artwork?

Of course not.

If you throw an AI algorithm at a blank screen, or at your starting image, or your idea, or some random factor that the AI uses as its starting point, the AI is no more the owner of the art than the paint bucket.

The AI has no intent, no purpose.

Even if the artist’s only apparent action is to throw the bucket — or to press a button — there is still the moment at the end where the artist has to decide if the result is worthwhile.

It is that simple act of decision that turns random splatter into art, and that turns AI-generated images into someone’s personal expression.

Until AIs become self-aware, that’s not going to go away.

If the AIs do become self-aware, and develop individual personalities, and motivations, and personal stories, and we start to care about them the way we care about people — well then, for all intents and purposes, they are people.