If your content being ingested by AI companies is the concern, just keep in mind that no copyright license will protect you as it doesn't cover this usage.

Considering there's no framework in place for AI companies to reasonably divulge how they trained their LLMs, assume, because Lemmy is on the surface level internet, that your comments have already been scraped and used to train every LLM in development.

IANAL, of course.

My understanding is that there is no such thing as an assumed licence for copyrighted work. The default is that there is no license and the owner reserves all rights. A license is a way to assign some or all of those rights to another entity.

The downside to copyleft is that you are giving up some of the ability to control the work, and it becomes harder to monetize as some people will question why they should pay for something they can have for free.

As far as I know, you can't use copyright to prevent someone from reading something that has been legally distributed to them. There are also some exceptions your government includes in copyright, like you can be quoted for the purposes of commentary and criticism, there may be exceptions for copies made for educational purposes or an exception that allows the government to make a copy for their archives, or if someone has legal access they might be able to make copies for personal use.

In all cases check your local laws. Copy left is a way to exploit these laws to allow humans to more easily benefit from your work but keep exploitative entities away. All of these only work within a legal framework however.