Such a one-sided argument, There was a summer of BLM riots with way more violence going on for months, dozens killed , hundreds injured, countless shops looted and destroyed but that’s obviously “peaceful protests”.
The whole study is an like a hammer looking for a nail, there is no evidence for anything like backprop, “credit assignment” or reinforcement learning in the human or other animal brain.
This would require some kind of global “teaching signal” and iterative optimization which is not biologically plausible.