when social media AI algorithms facilitate hate crimes, as discussed in previous chapters, who should be held liable?