Hero Image

The dilemmas of AI in creative domains like art and writing

Artificial intelligence has been used to help explore new disease treatments, to help product teams analyze and pull obscure insights out of the data they've compiled, and more. It can be an incredibly useful tool. I've written in the past about how when I was sick with COVID I explored creating a game with ChatGPT's help. While it was fun to ask the robot pal, it ultimately didn't sit right with me. As an artist, trained illustrator, and designer, I know how hard it can be to get creative work. I also know how enjoyable and emotional creating that type of work can be, but also how important it can be that it's informed by the culture around us. Things AI should stay out of.

A few questions

  • What happens if there's no longer new human creation to train AI on?
  • What kind reductive hell happens when AI starts training on its own creations?
  • How does AI create relevant material without stealing from artists and other creators?
  • Who truly owns the creation of an AI if it has been trained on the work of others, often without their consent?
  • As our culture changes, and our values as a society mature, how does AI evolve with us?
  • What's the danger to society of AI staying reductive and not adding new value and commentary?
  • Why use AI to create and infringe in areas where creatives are already having trouble getting consistent work, pay, and rights?

I hope that even posing these questions causes your brain to spin like mine. I'll pepper some more mind-spinning questions throughout, below. The way AI companies blatantly infringe on copywritten work and creators' unique voices to train their models on says a lot about how much we understand and value the arts and their contribution to society. Throughout history, entire art styles and techniques have been created and used just to push on the cultural zeitgeist. The loss of human touch and emotion in creative works is a real risk with these types of advancements.

Devaluing human creativity and craftsmanship

Who should receive credit and compensation for AI-generated creative works? There currently doesn't exist any Spotify-esque model that's compensating original creators the more you use an AI tool to reference their work. AI tools may facilitate the creation process, but human input is required to train and guide AI systems. Without proper attribution, compensation, and moral paramters in place all we've arrived at is exploitation. Exploitation that only further diminishes how creative contributions from artists, designers, illustrators, writers, and others are perceived.

Somewhat ironically, it's not just the creators that AI is being trained on that are being exploited, but also the very developers of AI technology in some cases. Noēma broke a story on this very topic, where they reported "So-called AI systems are fueled by millions of underpaid workers around the world, performing repetitive tasks under precarious labor conditions." And SFU's student-led paper reported "as little as $1.46/hour after tax." It seems like the entire point of AI is profit and exploitation. Sadly, that's not the only type of exploitation being reported by SFU, Noēma, and others. With AI datasets coming from an imperfect public, the AIs using them have been creating sexualized images, perpetuating things like harmful stereotypes, racism, infantilization, violence, and more. They spread our hate.

  • Why don't we compensate creators whose work is being used to train AI?
  • Why don't we fairly compensate the developers and other workers who create AI technology?
  • Why are we allowing AI to perpetuate and fuel the worst parts of us?
  • Who is held responsible when AI technology fuels hate crimes?
  • Why aren't AI companies more transparent, accountable, fair, and inclusive? (It's always about money, isn't it?)

What happens to the creators?

We have the devaluation of human creativity and craftsmanship occurring, and the exploitation of both creators and developers. When we start to displace creators from their own fields, what happens? I think there's going to be a tremendous effect on mental health and well-being of creators, for starters. AI will also take the jobs of creators, or perpetuate the existing power dynamic problems within creative industries. Since we've allowed AI companies to approach things in such a manner we'll have also opened the doors for the same thing to happen to other industries.

Displacing creators and causing them further problems isn't all we have to contend with, though. Where does all the money go? The money that would usually go to adequately compensating creators is instead being concetrated in AI corporations– most notable, their leadership, of course. Because why pay anyone involved in exploitation fairly?

  • We know money flows up, but does so much of it have to flow up?
  • Why can't creators be compensated if AI companies and their users benefit from their work?
  • What supports are being provided for creators displaced by AI?

What about the environment?

I encourage folks to check out even just this one article from Earth.org. They outline so many facets of this topic, like the chips AI servers require, the water usage of the servers, the carbon dioxide output, and more.

Can we balance advancement with ethics?

Many will argue that we need innovation and advancement and that these points and questions are merely roadblocks they have to break down in order to plough forward. We've started to see some creative communities like DeviantArt and Tumblr selling their users' data and creative work– allowing us to opt out if we don't want to take part. Money and poor ethics on their part might be why organizations like theirs can't instead use opt-in and compensate users' who want to do so. If you want to opt into helping train AI, all the power to you, but where is the compensation, and what about those who can't opt out? Many people have forgotten their accounts or have even passed away– what happens to their work when they don't opt-out? These companies and their beneficiaries don't care.

I think we can do better while still building for the future.

A few things AI companies could do better:

  1. Be transparent about whose creative work was used to inform AI creations.
  2. Invite people to allow their works to be used to train AI.
  3. Avoid using work from those who have not opted in.
  4. Compensate creators whenever their work has informed an AI creation.
  5. Ensure hate and prejudice isn't perpetuated by AI.
  6. Support creators displaced by AI.
  7. Continually evaluate and adapt AI against emerging ethical frameworks.

Which, of course, makes me wonder things like...

  • Why can't we use opt-in methods instead of opt-out for sharing user data and creations?
  • Why aren't creative communities like DeviantArt and Tumblr compensating creators after the sale of their information and works?

What can you, we, and me do?

I, personally, pledge that I will not use any exploitive AI tools to create games and assets for Monkey's Lunch. Items mentioned in my previous post will be scrapped entirely and revisitted.

I'm no lawyer or anything, but I have been slowly writing a book on the ethical creation of products and services, which has helped me create and pledge the below. I encourage you to take the following pledge with me, in whole or in part:

AI Abstinence Pledge

While I recognize the near impossibility to avoid AI technology in every setting, as a conscientious individual, I recognize the profound ethical, societal, and environmental implications of artificial intelligence (AI) technologies. In alignment with my values and commitment to ethical conduct, I pledge to refrain from using or supporting exploitive AI, including but not limited to:

  1. Use: I will not advocate for or participate in the implementation of AI technologies in professional settings where their deployment may compromise ethical principles, human rights, or the well-being of individuals. If I have used AI in creation of my work, in whole or in part, I will be transparent about it.
  2. Support and investment: I will refrain from financially supporting companies and organizations involved in the development, deployment, or promotion of AI technologies that lack transparency, accountability, compensation for creators, or adherence to ethical standards.
  3. Community advocacy: I will support non-AI creators, their rights, compensation, and job security.
  4. Advocacy and awareness: I will actively raise awareness about the ethical, social, and environmental implications of AI technologies, engaging in constructive dialogue and advocacy efforts to promote responsible AI development and deployment.
  5. Educational initiatives: I will support educational initiatives aimed at fostering critical thinking, digital literacy, and ethical awareness among individuals, particularly regarding the implications of AI technologies.
  6. Environmental considerations: I will consider the environmental impact of AI technologies, including their energy consumption and carbon footprint, and advocate for sustainable alternatives that prioritize environmental stewardship.
  7. Community engagement: I will engage with my communities to promote dialogue, collaboration, and collective action on issues related to AI ethics, accountability, and societal impact.
  8. Continuous learning: I will remain open to new information, perspectives, and developments in the field of AI ethics, committing to ongoing learning and reflection to inform my ethical stance and actions. I will push AI organizations, supporters, and enthusiasts to also partake in continuous learning and application in AI ethics and acceptable use.
  9. Accountability: I will take part in holding AI companies and those organizations they hire for data modelling and other use accountable, pushing them for transparency, environmental and other sustainability considerations, voluntary participation from creators, fair compensation, attribution for creators, and other ethical considerations.

By taking this pledge, I affirm my commitment to ethical conduct, social responsibility, and the well-being of present and future generations. I recognize that my individual actions contribute to shaping the ethical trajectory of AI technologies and their impact on society, and I pledge to stand against exploitive and unethical technologies in society at large.

Disclosure of use/non-use and creation

One thing I've found particularly interesting as a means of rebelling against AI companies or those who create work using them is to openly disclose when you've made something without AI. Folks like Hinokodo have provided asset packs of emblems, badges, or stamps to emblazon your work with and let people know "Made By Human." Of course, I would also prefer if those who use AI dislcosed things in a similar manner.

Hire creatives!

If you're working on something and you want help, please, please, please hire.

I've been trying to hire via UpWork and having trouble finding a good fit, but I've found the best way to find folks to work with to just be via social media. I recently hired Hodag to help with the art for Bug & Claw and it was so refreshing to see their take on the little post-apocalyptic world I've been creating. I plan on hiring them again, soon, once I have more budget alotted! I followed them for a long time on Twitter (I will not call it X, sorry), before reaching out and it was that easy.

I tried to hire someone via Fiverr after someone recommended it (I honestly thought it was only for things that cost $5, which I thought was super-exploitive, but that's not the case anymore). You'll never guess where I ran into someone using AI. So, fair warning, when trying to hire on sites like that it seems like we have to disclose now that we won't accept work made in whole or part by AI. Honestly kind of blew my mind that was a thing I had to define, but that's where we are.

An argument I keep hearing from people is "but I can't afford it right now." To which I reply: there are other options.

  • Seek quotes from multiple sources. New folks vs industry veterans charge different rates.
  • Heck, you can even try to find someone to split revenues with you if you can't afford it right now.
  • You can attempt to do what you need yourself.
  • If you aren't good enough yourself but the gap isn't that large you can ask for feedback across various social media and other sites and others can help you bring it hte last 10%.
  • Just don't include art/that portion of writing/whatever for now. Iterate on it later.
  • Be patient and extend the timeframe of your project so you can pay people over a longer period.
  • Use public domain sources like Unsplash.

Final thoughts

AI is exploitive, currently bad for the planet and society, and devalues creative people and their work, and there are other things we can do instead of using AI. We also need to hold AI companies and their patrons to account. Innovation and advancement are inevitable, but that doesn't mean they can't come with proper ethical consideration and while avoiding exploitation.