This article sounds more like a hit piece against AI rather than actually caring with a CSAM-free internet: 1) LAION only contains links. It's the same links o

This article sounds more like a hit piece against AI rather than actually caring with a CSAM-free internet: 1) LAION only contains links. It's the same links one would find on search engines, or any other large datasets. If they wanted to make this seem less biased, they could have framed as "Identifying and eliminating CSAM in public internet datasets", no need to even mention AI or Machine Learning. But I guess they really want to portray those datasets as some sort of "CSAM database" to tarnish AI reputation in the public eye... 2) You could write this same article about, for instance, search engine indexes or the Internet Archive. Which most likely host such material somewhere there as well. But it doesn't involve AI, so those researchers don't care. 3) The only reason why these links were found and remove is because it's a public dataset. Do they prefer that companies use private datasets, where no one knows what is in it, so no "researcher" can write their little misleading manipulative "think of the children!!" article on it?