With #AI, my main concern lies in things like predictive policing, facial recognition, hiring algorithms, landlord tenant algorithms, tax collection algorithms (yes, these have gone horribly wrong), moderation algorithms, and so on.

These sorts of algorithms where power is imposed by a powerful party onto a more vulnerable party (and which are inherently discriminative).

I'm not really concerned about generative "AI" (which is really a hyped up threat in a number of ways, it's up to someone to be responsible).