I have recently shared several AI-generated comics online about my experience of this year’s heatwave and about animal agriculture’s contribution to climate change.
Because I disclosed that the comics were AI-generated, however, the discussion quickly shifted away from the issues I wanted to raise. Instead, people began criticizing my use of AI image generation. Their objections included electricity consumption, water use, copyright concerns, factual and visual errors, poor artistic taste, and general “AI fatigue.”
I do feel a bit uncertain about the ethics of what I am doing. I have created many AI-generated comics, posters, and other images to promote veganism and climate awareness, but I do not know whether these materials have ultimately done more good than harm. AI companies are not transparent enough about the environmental cost of generating images, so it is possible that I should stop using these tools.
What troubles me even more is the strong emotional hostility toward AI-generated images. I can often make an image look much less obviously AI-generated. In informal experiments, when I asked people to identify which of two images was made with AI without signaling the answer, their guesses were often no better than random. But deliberately hiding the use of AI would cross an ethical line for me. At the same time, very clearly disclosing it can provoke strong reactions such as: “How can you use AI while claiming to care about ethics or the environment?” In practice, I usually do not mention AI unless someone asks, but I answer honestly when they do.
So how should activists think about this? Should they avoid AI-generated images altogether? More broadly, should they avoid any AI-generated content in public communication?
That said, many (most?) parts of society have a strong anti-AI preference, and strategically it doesn’t make sense to anger your audience.
I think it is worth separating “don’t do this because it’s morally wrong” from “don’t do this because it’s not effective for winning hearts and minds”.
For my part, I dislike AI images, and I feel disappointed when creators use them in blog posts, videos, and other places. I value human authorship and prefer to consume content that a person actually made.
At the same time, I think the opposition to generative AI is often misinformed and disproportionate to its actual impact, and I condemn the false accusations and creators getting cancelled for using AI in ways their audiences don’t like. Hank Green is a recent example of someone who used AI in a relatively minor way in a video and is still getting cancelled after he apologized to his audience for it.
At the end of the day, I think it’s up to creators to know what their audiences want. Feel free to use AI art if your audience is receptive to it, otherwise don’t. I also believe that the EA media ecosystem would benefit from more diverse content—maybe not every podcast should be a video podcast, and not every blog post should use AI images.
There are good strategic reasons to avoid AI art (because people get incensed). As an activist, one of your goals is to influence public perception. For many campaigns, using AI art is going to be net negative for your cause, because they’ll associate your cause with something they dislike.
I don’t agree with the object level reasons to avoid AI art, but that doesn’t change the strategic considerations.
The cartoon market is grim—there are a lot of cartoonists looking for projects and an outlet for their work. Find a creative way to partner up with one of them. Offer a token stipend or make it a submission contest (if you have enough of an audience, it’s a nice way to give them an audience).
Just don’t tell them your funny idea. If you are inclined to do that, get a pen and draw it yourself. There’s not as much risk of making “bad art” as you might think if the writing is good. If you have an idea, sketch it up—anything handmade that doesn’t hide its grungy authenticity, I think, is so preferable to even the best AI cartoons.
The limited range of AI cartoons (like most, I can spot them in a millisecond) can’t help but pull attention away from any message and focus it on how the cartoon was produced. But honestly, why would you expect it to be otherwise? You are using a format that is defined as the message conveyed by the combination of the image and the writing, so how the image was made is half the message. Some of the best cartoons (and they are the hardest to make) have no caption—the drawing is doing all of the narrative. They are insanely thought-through documents.
Consider this an unforced but predictable error; trying to pass a message via an endangered art form without consideration for how it was produced is going to get a reaction. Just like oil painting and sculpture, in cartooning, form and content are inseparable. That’s what makes it art instead of clipart.
I think the compute you spend on image generation is probably very little, your work is very likely net positive for the environment, and hiring a designer to make your images would be prohibitively expensive.
While that is likely true, audience perception matters, and you can’t logic your audience into liking AI images when they don’t.
Also, there are plenty of images on Wikimedia Commons—e.g. images of factory farms, chickens, and pigeons—that are free to use, usually with credit required. I’d reach for these if you don’t have the budget for an illustrator.
Thank you all for the suggestions. After thinking a bit more about it, I think I am going to:
Work mostly from my own photos and free stock images rather than generating from scratch.
Still use an AI agent for minor editing, collage, and text overlay on those images.
For posters, design the layout myself. Then generate a simple background based on that design. Use agents to add text overlay, QR code etc to finish.
After some digging, I do agree that AI image generation’s environmental impact is much less than, say, eating a burger. Still, I’ll generate fewer variants than I used to. It’s almost always good to be frugal in terms of consumption!
[Question] Should we avoid using AI images in activism?
I have recently shared several AI-generated comics online about my experience of this year’s heatwave and about animal agriculture’s contribution to climate change.
Because I disclosed that the comics were AI-generated, however, the discussion quickly shifted away from the issues I wanted to raise. Instead, people began criticizing my use of AI image generation. Their objections included electricity consumption, water use, copyright concerns, factual and visual errors, poor artistic taste, and general “AI fatigue.”
I do feel a bit uncertain about the ethics of what I am doing. I have created many AI-generated comics, posters, and other images to promote veganism and climate awareness, but I do not know whether these materials have ultimately done more good than harm. AI companies are not transparent enough about the environmental cost of generating images, so it is possible that I should stop using these tools.
What troubles me even more is the strong emotional hostility toward AI-generated images. I can often make an image look much less obviously AI-generated. In informal experiments, when I asked people to identify which of two images was made with AI without signaling the answer, their guesses were often no better than random. But deliberately hiding the use of AI would cross an ethical line for me. At the same time, very clearly disclosing it can provoke strong reactions such as: “How can you use AI while claiming to care about ethics or the environment?” In practice, I usually do not mention AI unless someone asks, but I answer honestly when they do.
So how should activists think about this? Should they avoid AI-generated images altogether? More broadly, should they avoid any AI-generated content in public communication?
I wouldn’t worry much about this, those costs on a per-user basis are incredibly low, almost certainly much lower than other things you won’t worry about. Andy Masley is a great source on the topic: https://blog.andymasley.com/p/a-cheat-sheet-for-conversations-about
That said, many (most?) parts of society have a strong anti-AI preference, and strategically it doesn’t make sense to anger your audience.
I think it is worth separating “don’t do this because it’s morally wrong” from “don’t do this because it’s not effective for winning hearts and minds”.
I think the post “Debate: should EA avoid using AI art outside of research?” from April 2025 is relevant here.
For my part, I dislike AI images, and I feel disappointed when creators use them in blog posts, videos, and other places. I value human authorship and prefer to consume content that a person actually made.
At the same time, I think the opposition to generative AI is often misinformed and disproportionate to its actual impact, and I condemn the false accusations and creators getting cancelled for using AI in ways their audiences don’t like. Hank Green is a recent example of someone who used AI in a relatively minor way in a video and is still getting cancelled after he apologized to his audience for it.
At the end of the day, I think it’s up to creators to know what their audiences want. Feel free to use AI art if your audience is receptive to it, otherwise don’t. I also believe that the EA media ecosystem would benefit from more diverse content—maybe not every podcast should be a video podcast, and not every blog post should use AI images.
There are good strategic reasons to avoid AI art (because people get incensed). As an activist, one of your goals is to influence public perception. For many campaigns, using AI art is going to be net negative for your cause, because they’ll associate your cause with something they dislike.
I don’t agree with the object level reasons to avoid AI art, but that doesn’t change the strategic considerations.
A Cartoonist’s Perspective
The cartoon market is grim—there are a lot of cartoonists looking for projects and an outlet for their work. Find a creative way to partner up with one of them. Offer a token stipend or make it a submission contest (if you have enough of an audience, it’s a nice way to give them an audience).
Just don’t tell them your funny idea. If you are inclined to do that, get a pen and draw it yourself. There’s not as much risk of making “bad art” as you might think if the writing is good. If you have an idea, sketch it up—anything handmade that doesn’t hide its grungy authenticity, I think, is so preferable to even the best AI cartoons.
The limited range of AI cartoons (like most, I can spot them in a millisecond) can’t help but pull attention away from any message and focus it on how the cartoon was produced. But honestly, why would you expect it to be otherwise? You are using a format that is defined as the message conveyed by the combination of the image and the writing, so how the image was made is half the message. Some of the best cartoons (and they are the hardest to make) have no caption—the drawing is doing all of the narrative. They are insanely thought-through documents.
Consider this an unforced but predictable error; trying to pass a message via an endangered art form without consideration for how it was produced is going to get a reaction. Just like oil painting and sculpture, in cartooning, form and content are inseparable. That’s what makes it art instead of clipart.
I think the compute you spend on image generation is probably very little, your work is very likely net positive for the environment, and hiring a designer to make your images would be prohibitively expensive.
While that is likely true, audience perception matters, and you can’t logic your audience into liking AI images when they don’t.
Also, there are plenty of images on Wikimedia Commons—e.g. images of factory farms, chickens, and pigeons—that are free to use, usually with credit required. I’d reach for these if you don’t have the budget for an illustrator.
Thank you all for the suggestions. After thinking a bit more about it, I think I am going to:
Work mostly from my own photos and free stock images rather than generating from scratch.
Still use an AI agent for minor editing, collage, and text overlay on those images.
For posters, design the layout myself. Then generate a simple background based on that design. Use agents to add text overlay, QR code etc to finish.
After some digging, I do agree that AI image generation’s environmental impact is much less than, say, eating a burger. Still, I’ll generate fewer variants than I used to. It’s almost always good to be frugal in terms of consumption!
Here are some sources I found:
https://morethanmeatstheeye.substack.com/p/ai-water-use-vs-animal-agriculture
https://blog.andymasley.com/p/individual-ai-use-is-not-bad-for
https://news.mit.edu/2025/explained-generative-ai-environmental-impact-0117