> Does this translate to "Applications of this research might destroy society, please refrain"?

This is boilerplate acknowledgement of ethics concerns that often appears in DL/ML papers these days. Doesn't really mean anything.