AI might become the artist's best friend rather than making them obsolete in the future. Here we explore some ways in which AI is actually helping improve artists' lives rather than threaten their livelihoods.
Can a computer be creative?
Did you know that computers are incredibly stupid? They are not able to diverge from written code and will follow them blindly to their ultimate conclusion.
Despite their ever-growing ability to perform calculations in a fraction of a second, you would have a hard time having a meaningful conversation with one. Though there have been some interesting examples of computers appearing smart in the past.
"BM’s Watson supercomputer defeated two top Jeopardy! players last year, but for the clue “What grasshoppers eat,” Watson answered: “Kosher.” For all the data he could access within a fraction of a second—one of the greatest corpora ever assembled—Watson looked awfully dumb." - nplusonemag.com.
Beyond natural language understanding, and others are computers appear to be clever is through AI. AI-art is a prime example.
It gives the appearance of creativity but "all that glitters is not gold". Behind the mighty algorithms that make up an AI, there is a team of human coders effectively giving it the instructions to be "creative".
But are humans any different at the end of the day? We are also effectively soft machines whose bodies and minds run on a series of basic rules set by your genome. We are effectively programmed using the base pairs of DNA rather than the binary of computers.
AI-artist is effectively the same and follows a strict set of rules, or defined variables, to create its final creative piece. While some can write their own code, the way in which it can do this has been pre-defined by its human coders in the first place.
But it should be noted that machine-learning and neural networks can learn from their past mistakes and experiences. In this sense, they are pretty analogous to human beings.
While the results of AI creativity can be unpredictable, is this really creative? Or just an error in the code? We'll let you decide.
How AI is helping artists, not threatening them
It turns out that artists have very little to fear from AI supplanting them in the future. Here are some reasons why.
1. AI tools help save them time
Many existing software packages used by artists, like Adobe's suite of services, are starting to incorporate some basic AI's to help automate basic repetitive tasks. The hope is to help speed up the process of using their tools and reduce the need for time-consuming manual processes.
These include things like machine learning tools to help find specific video frames, to other features that color in entire works of line art with just the click of a button. These kinds of functions are nothing really, with software packages like Adobe Firework's "Magic Wand" feature, for example.
This basic tool automatically works out which parts of an image to select to cut, paste, fill or delete. For other AI-assisted graphic design packages, features like smart cropping or automatic photo tagging to enable an artist's work to be discovered online.
Another interesting area of development is auto-coloring. While this might sound like cheating, it should help artists save a ton of time. primarily designed for comics and animations, Celsys' Clip Studio now includes just this kind of feature.
With some basic instructions from the artists, the AI takes over to turn black and white line drawings into full masterpieces. Results can be unpredictable and may need some human cleanup, but its potential is huge.
2. AI actually helps artists be more creative
There are only 24 hours in a day, and you spend a big chunk of that time sleeping and eating. So, any assistance AI can offer an artist, like handling repetitive tasks, will significantly help the artists be more creative.
The importance of this can't be understated. With more time on their hands, artists can spend more time honing their craft.
As AI and ML digital art package features get more sophisticated, it also adds new tools for artists to play around within creative and unpredictable ways. AI-solutions can also be used to create things beyond what an artist might have ever thought was possible before.
Especially when compared to more traditional, labor-intensive methods like oil painting, for example. Just what will be possible in the future is anyone's guess, but it is quite exciting to dream of the possibilities.
3. Imitation is the best form of flattery
One major area that AI is helping creatives with their craft, is by helping them imitate the styles of famous artists from the past. The Runway ML hub, for example, lets people upload photos and have an AI render them in the style of long-dead painters.
But it's not all just about imitating great artists, the software can also help create more unique AI-assisted artwork. From turning text into images to helping render doodles into landscapes, AI-tools like this are added extra weapons to an artist's arsenal.
4. Creative projects are becoming more collaborative helping cut project completion
AI is helping clients, managers and artists collaborate more easily than ever before. In the past, a project brief was agreed and each member of the chain would head off into their own corner to get their bit done.
Now with advents in AI-powered innovations like Agile and Google's Spirit, it helps the "ideation" process open up to all stakeholders. This will help speed up project completion and ensure everyone is aware of the final objectives of the creative project in question.
5. AI is helping forge a new genre of art
These neural networks make use of a generative adversarial network (GAN) and the products of these systems in both spooky and fascinating at the same time. Some of the more famous pieces, like "Portrait of Edmond Belamy" were recently sold at an auction for an astonishing $432,500.
The success of pieces like the one mentioned above has led some to believe we are on the cusp of a "Gold Rush" of AI-art in the near future.
While this might sound like the time of the human artists is at an end, the truth is far more nuanced. The code behind these GAN's, as well as the set parameters for the goal are all defined by humans in the first place.
In essence, the idea behind a piece is still within the domain of the human mind. The AI (GAN) simply 'does the leg work' on their behalf.
The potential for this in the future is practically limitless. Who knows what Picasso could have achieved with such a GAN.