It means that people should not be surprised that AI companies are training on their data. They're deliberately putting their content out into the world where AI trainers can read it in an uncontrolled manner, and reading it is all that's needed for AI training.
There have already been a number of lawsuits about AI training and thus far nothing seems to indicate that it's something that copyright restricts. If you know of any cases that have established otherwise I suppose feel free to link them, but until then there's nothing illegal going on here.
If you just want to be angry about it then I suppose there's nothing stopping you on that count. Go ahead.
