Summary:
I downvoted pro-AI comments in a post in leftymemes community. It was LLM generated polandball comic (Which is objectively pathetic as fuck) that showed up on my feed, blocked couple of users who I thought were unhinged, and have blocked the whole instance on my client after realizing how rabid these morons are.
I didn't go looking for AI posts like a vigilante.
One user in question got miffed for being downvoted and banned me from places they moderate.
https://www.databricks.com/blog/stable-diffusion-2
AI model training is getting more efficient over time, not less. And each model only needs to be trained once, so your analogy doesn't quite work.
https://huggingface.co/docs/diffusers/main/en/tutorials/basic_training
You can absolutely train a small, specialized model on minimal hardware.
It's also locally hosted and free to use, so each image generated does not directly incentivize the training of future models, unlike cloud-based subscription services.
Worrying about open-source AI while corporate AI burns orders of magnitude more resources every day is much like everything else with climate change. Even if you somehow stopped it, it would barely be a drop in the bucket against the massive, inefficient behemoths destroying the world.