ADHD memes

I absolutely can deny that it's unethical. In fact, you could argue that using an LLM for memes is not only morally neutral, but actually morally positive.

First, environmental concerns. Is there any evidence that it takes more energy to quickly generate a meme image than to draw it on a computer? That seems highly unlikely. You would have to power a computer for hours to produce such a drawing by hand. It almost certainly takes a tiny fraction of the energy to generate an image via an LLM than it would to generate it via computer drawing software. And even hand-drawing doesn't escape this, as all the supplies you need also have high energy input requirements.

As to your other general ethics point, that's superstitious mumbo jumbo. You're arguing that an LLM is magically cursed because of how it was created. A tool can be created ethically or unethically, but something being created unethically doesn't magically make its use unethical. The only potential issue is that if you're funding the creation of the tool or giving the toolmaker money, you can provide a greater market for the creation of more unethical tools.

If anything, using the LLMs of the big AI companies to make memes is an absolute moral good. No one is going to pay OpenAI to make memes. Meme images cost the AI companies money to make but don't provide them any revenue at all.

Every meme image made by AI is a small blow against these companies. It costs them money but generates them no revenue. But because apparently AI is magically cursed with evil, any use of it must also be evil.

This is religious thinking, not rational thinking.