This post is part of Lifehacker’s “Living With AI” series: we explore the current state of artificial intelligence, discuss how it can be useful (and how it can’t), and assess where this revolutionary technology is heading. Read more here.
AI-generated art is not a concept: it is here. Thanks to numerous tools with a elementary and accessible interface, anyone can jump on their computer and start generating any image ideas that come to their mind. However, as more and more people have started experimenting with these tools, grave ethical and legal issues have arisen, and it seems that almost everyone on the Internet has an opinion on this divisive technology.
As part of our series on living with AI, we’ve put together this guide to explain how AI art tools work, explain the controversies surrounding them, and show how they impact everyone from professional artists to the curious.
Where and how to create AI art
Before we delve too deeply into the technology and ethics of AI art, let’s quickly look at the tools themselves.
There are many AI art generators, but the main players will be Halfway through the journey, Stable diffusion, Bing AI Image Maker, OD-2, CrayonAND Dream. All of these tools are available via the web or computer, and Dream also has a mobile app.
Midjourney is considered the most powerful, but requires a subscription starting at $10 per month. Midjourney also requires a Discord account, as it runs entirely through a dedicated Discord chat server. This means you collaborate with other users and all your photos are published publicly on Midjourney’s online gallery, unless you pay $60 a month for the “hidden mode” feature.
DALLE-2 also produces high-quality images, but fresh users must pay at least $15 to register and operate it. Although your session and creations are private, it uses a confusing token system where each word in the prompt costs a certain amount, and if you exceed your token limit, you have to pay extra.
The Bing AI Generator and Stable Diffusion are completely free and allow you to create as many images as you want, but images take longer to generate, especially if the servers are busy. Craiyon is also free, but requires longer generation time and lower image quality. Finally, Dream offers a decent compromise between quality and price, allowing you to make one image at a time for free, or up to four with a paid account starting at $10 per month.
In terms of affordability, the Bing AI Generator is definitely the best option if you’re simply interested in these tools. It’s free and can be accessed from the Bing Search homepage or in the Edge browser.
That said, despite differences in quality and interface, all of these tools work the same way: type a prompt in the text box describing the image you want to see, press Enter, and then wait a few moments for the AI to generate an image based on your description.
The quality of the final product will depend on the tool used and the detail of the prompts. Some tools, such as Midjourney and Stable Diffusion, provide ideation tips better hintsand even additional features that make this possible aid the AI get closer to the intended result. However, even with these additional steps, the process only takes a few moments and each tool is elementary to operate.
How do these tools teach you to draw? And why are there too many fingers?
The images produced by these tools may be impressive, but not because the software can actually draw.
As I mentioned earlier this year, calling these products “AI” is misleading. Contrary to the common concept of artificial intelligence presented in science fiction media, these tools are not living, sentient or in any way conscious, they do not reason or learn. This applies to both text-based chatbots like ChatGPT and generative graphics tools like Midjourney. Simply put, they work like predictive text on your phone, pulling in a list of possible solutions for suggestions. When it comes to generative graphics tools specifically, the tool simply searches for images that match keywords or descriptions in the prompt and then puts the pieces together.
This is completely different from the actual drawing process – the AI never actually “draws” anything, which is why these tools are eminent for “can’t draw hands.”
As video game designer Doc Burford explain, “If I tell a machine, ‘show me Nic Cage dressed as Superman,’ the machine may have images labeled ‘Nic Cage,’ and it may have images labeled ‘Superman,’ but wherever the thinking mind of true intelligence places them, we combine ideas and fill in the blanks.” places with things it knows – like an artist who has also memorized human anatomy – the AI will still give me an imperfect S shield on Superman’s chest, which will mess up my fingers.”
Ethical and legal issues of AI art
These tools are straightforward to operate and can often produce convincing results – apart from the occasional extra digits and weird faces – but there are grave ethical concerns about the creation and distribution of AI-generated art that go beyond quality and accuracy.
The main problem with AI generative art tools is that they rely on the names of uncredited, unpaid artists whose works are used without consent. Each generated image exists only because of the artists from whom it is copied, even if those works are not copyrighted. Some AI evangelists like to claim that these tools work “just like the human brain” and that “human artists draw inspiration from or reference other artists in the same way,” but this is untrue for many reasons.
First, artificial intelligence has no essence or mind, and therefore no memory, intention, or skill. Stable Diffusion isn’t about “learning” how to draw or getting inspiration from another piece of art: it’s just an algorithm that finds and automatically populates data the way it’s programmed. On the other hand, people think, feel and act consciously. Their works result from memorized skills and life experience. Even using someone else’s work as a reference or inspiration is a conscious choice driven by the artist’s goals.
To aid clarify this distinction, I reached out to Nicholas Kole, an illustrator and character designer who works with major film and video game studios such as Disney, Activision, and Dreamworks. “The work I do as a concept artist and illustrator begins with a deep dive into the context of each project,” he says. “I ask specific questions and give ideas about world building, story and gameplay [and] the process from start to finish is extremely specific, tailored and tailored to the specific needs of my colleagues and clients. Every cufflink, belt buckle, prop and motif – the art we create is a careful work of thoughtful design, crafted with love and attention to detail.”
“To have an algorithm intrude into this process that doesn’t care about context, that doesn’t know whether people have 5 or 17 fingers, that puts together visual guesses based on stolen data, and basically works like a million monkeys at once with a million typewriters is anathema to me.” .
Kole says the art of artificial intelligence “goes against everything I represent creatively and everything I wanted to do in my life. This is an insult to the reason I create and engage in art – I want to see thoughtful human craftsmanship that is expressive, and to express myself through thoughtful human craftsmanship.” A quick glance at portfolio sites like Art Station shows that Kole is not alone in these views, as many professional artists are staunchly opposed to AI art.
This tough stance is not solely due to ideological or aesthetic reasons. AI automation is posing a threat to job security in many industries — which is why unionized actors, screenwriters, Google contractors, and even airline workers are currently on strike across America. The threat to working artists is equally real.
AI art also poses risks to the companies that employ these artists. There have already been significant legal disputes over AI artworks infringing copyrighted material, and the current results favor the original artists. Such as, many companies ban it completely operate of AI art and reject any submissions from artists who have AI-generated works in their portfolio to avoid copyright issues.
Are there ethical applications of artificial intelligence art?
Despite the ethical and legal issues, some argue that these tools have a place and can be helpful even for professional artists. IN interview with KotakuVisual artist RJ Palmers says artists could operate AI to “create loose compositions, color patterns, lighting, etc.” and that the tools “can be very nice for gaining inspiration.”
Likewise the author and animator Scott Sullivan argues on his blog that AI is helpful in generating ideas and iterating during brainstorming, and that “it all depends on the intention of the artist and how he uses the tool.”
But while AI art is controversial among professional artists, regular users may wonder whether it makes any difference to the layman or hobbyist who just wants to have fun with them every now and then. And of course, AI tools can be used as a toy, but this should be kept in mind NO how the creators of these products treat them.
Almost all AI graphics generators are commercial products in some way. Some of them are paid products like Midjourney or DALLE-2, while free services like the Bing AI image generator make Microsoft money through advertising revenue. Some of them also serve as “proof of concept” examples to encourage commercial customers to pay for a more powerful version of the tool.
In all cases, the people creating these tools make money from the work of the artists whose work is used to create the image you generate, even if it’s just for fun. As Kole explains: “The generative system cannot function without the labor of the stolen lives of countless enthusiasts like me. Each of them brought their lived experiences, opinions, fixations and points of view to their works, and now unexpertly combine them into one whole and advertise them as original works of art.” Even if you don’t share or sell the images you create, many of these tools maintain a public record of all generated content that other users can download and distribute.
Given all these concerns, it’s tough to recommend AI art creators, even if the intent behind their operate is innocent. However, these tools already exist and unless future regulations force them to change, we can’t stop people from trying them. However, if you do so, please be aware of the legal and ethical issues surrounding creating and sharing AI art, think twice before sharing it, and never consider an AI-generated image to be your own work.
