I'm not quite sure it's as simple or innocent as saying that using AI images are the same as bait and switch tactics.
LLM generative AI models are flagrantly stealing people's likeness by scraping the internet for photos (both SFW and NSFW) and then training itself on that content to create references for what it believes "people" look like both clothed and unclothed. When someone then prompts an LLM AI system to create an image of themself this undeniably happens at the expense of everyone whose photos were stolen and exploited to train the AI. The greatest danger here, and it's not even a hypothetical, is that sexually explicit images can be created without the consent of the people being referenced when getting generated by these AI photos. It is a known fact that LLM AI models are peverting the images it scrapes and allowing users to generate abusive material depicting children, let alone the non-consensual images of adults whose photos were stolen in order to train AI on how to predict what human bodies look like naked. Whenever I see another SW using generative AI to make their "images" I just can't help but think they're going to end up in a special place of hell for being complicit in feeding the AI even more sexually suggestive (and sometimes entirely sexually explicit) content which is then used by the AI to further train its models on how to generate even more non-consensual porn.
All of this is to say nothing of the catastrophic environmental impacts of these AI data centres and of the horrible practices these companies shamelessly commit. For the most popular LLM AI companies, it's also not a secret that many of these CEOs and others in their C-suites are investing their profits with categorically facist intentions.
It's an absolute tragedy that instead of using AI technologies to find cures to diseases or discover innovations in agriculture to reduce food insecurity or develop new chemical compounds to create more affordable medications, we are stuck in this ridiculous dystopia where people are absurdly deciding to create fake pictures of theselves when they could just use their camera and actually take the picture??? I will never understand how people can be confronted with the option to either A) take a picture of themself, that's it; or B) willfully ask a computer to use stolen images to hallucinate an approximated depiction of themself, while simultaneously contributing to the training of an LLM generative AI model on how to create more realistic child sexual abuse material... and they still choose option B.
TLDR; yes, it bothers me.
