The Problem With AI Art Generating Tools

Posted: October 31st, 2022 ˑ Filled under: Articles, Technology ˑ  1 Comment

As everything in life, any new solution or technology invention never comes at a free cost. Are the ethics of using this technology subjective? Who gets to make that decision for the public? Even if the public decided not to use it, how do you fight the big brands already investing and using AI/Machine learning tools? This cat can never be put back in the bag; things in the art/graphics world will never be the same.

There are already a lot of posts explaining what an AI visual art system does, “a text-to-image program” but if you’re not very familiarized with it, this video might help:

I could be wrong, but I think that people who were not previously paying for art directly created by real life artists might be some of the most frequent consumers and creators of AI art. This right here is a controversy because many of these consumers were not putting their money into this because of its cost. So in theory, these tools provide access to almost everyone with a device and internet.

The subjectiveness of AI Art ethics starts with what the majority decides is acceptable as art. What is art? After how much effort would someone’s creation be considered “a work of art”. Though drawing, painting, sketching, photographing, are all skills that can be learned, there is no doubt there are people who are born with natural ability versus those who will be required to spend years of their lives to reach that level. In this scenario are both people the same level of artist, or does sweat and tears mean more than the end result? These are difficult questions and I’m trying to be as objective as possible, regardless of the ‘Silicon Valley’ tone the AI Art conversation is taking in many spaces.

These are some test results from the three programs I tested using this prompt: “4:7, hd, 8k, what is art, detailed, unreal engine, a cathedral by studio ghibli.”

Ghibli Cathedral Stable Diffusion

Stable Diffusion

Ghibli Cathedral Midjourney

Midjourney

Ghibli Cathedral DALL-E

DALL-E

It makes me curious to think, would this shift designers and commercial artists projects to be more about curating and editing like many book editors do for writers?

AI generated art is not free, rarely anything ever is. There are free trials, but sooner or later you have to have a paid subscription. DALL-E allows you limited monthly credits for now. Before you invest a lot of time testing any of these program, I suggest you to read the small print on their Terms and Conditions so that you can have a clear understanding of their copyright and legal uses.

I didn’t dive too much into their copyrights because all the art I created was to test for articles or to use as an image for a food article again for testing purposes. However, if ownership is something that you’re concerned with, do read their Terms of Service.

There are a lot of mixed feelings about the topic and I plan on interviewing an online artist friend whose work I have previously purchased. I love her work. I wonder if she will have a different take on what other artists have said about AI Art.

Joanna Penn from The Creative Penn had an insightful podcast episode on this topic who actually catapulted me to go for it and write this article. Her podcast episode goes over these programs and also how this can be helpful to authors.

I’ve been playing with an AI program for a few years now, This Person Doesn’t Exist. I have used it for inspiration on characters for different stories I’ve plotted or written. We can also talk about “deep fakes” and photo/video manipulation. The subject of image manipulation in general is always controversial, and it will only get better/worse, depending on how you feel about machine learning in general.

I posted a poll on Twitter, but I don’t have a big following so it probably won’t receive too many votes. Either way here are some tweets that caught my attention on the topic:

 

 


My tweet poll:

Cat Graffam, an artist YouTuber, shared their thoughts on AI Art and the issues it poses to all artists. I like to listen to what artists across all platforms have to say. I have some friends I asked about it. One, a photographer, said he didn’t know what AI art meant. So I think 2023 is going to get pretty crazy when it comes to this whole topic.

Not all artists react negatively to it, some are just amazed by its results, this was entertaining:

On Midjpurney I tested prompts in Spanish to see if it would provide proper results. I might revisit this topic and the real accessibility of these programs in a few months. Let’s see how much Spanish they can handle.

LittleLadyCook Midjourney Spanish Prompts

If I were to spend money to use any of these programs, it would be to test it for a graphic novel. However I can’t see finishing such project with at least consulting a human graphic artist. You can feed these AI programs all the prompts you can think of and advanced techniques and styles, but so far these machines don’t fight back against your ideas. So far that’s something you can only get from another human and that kind of feedback is priceless.

—

Other interesting articles on this topic:

Share
Tags: , , , , ,


One Comment

Leave a Reply